Buried in the federal charging rule is a number that decides more about interstate architecture than any target for vehicle sales. A charging station built with federal highway money along a designated corridor must have at least four network-connected direct current fast charging ports, each rated for continuous delivery of at least 150 kilowatts and each able to serve a vehicle’s request up to 150 kilowatts simultaneously with the others. Four times 150 is 600, and 600 kilowatts is a service size, not an accessory.
Commenters saw it coming. During the rulemaking, several argued for a station capability of 450 kilowatts rather than 600, on the reasoning that vehicles draw their highest power at the start of a session and four cars would rarely be at peak together. FHWA kept the simultaneous requirement, on the ground that a customer’s experience should not depend on which port they pick or how many others are busy, and added an explicit allowance for power sharing above the 150 kilowatt per-port floor.
What the regulation actually specifies
The standards were published on 28 February 2023 and are codified at 23 CFR part 680. Each fast charging port must support output voltages from 250 to 920 volts direct current and must carry at least one permanently attached CCS Type 1 connector, with other non-proprietary connectors allowed alongside it so long as every port can charge a CCS-compliant vehicle. Corridor stations must be available and physically accessible to the public 24 hours a day, seven days a week, year-round. Each charging port must achieve an average annual uptime greater than 97 percent, computed monthly across the previous twelve months.
An uptime obligation stated that precisely is the most consequential clause in the rule, because it converts a construction grant into an operating commitment. Three percent of a year is about eleven days. A port that goes dark for a transformer replacement, a payment terminal fault and a winter communications outage has spent its allowance.
The 2025 guidance moved the geography and left the engineering alone
A build-out figure for this program carries a date or it carries nothing, because the program was rewritten in the middle of its own construction. FHWA’s interim final guidance of 11 August 2025 supersedes all previous guidance, naming the June 2024 program guidance and the December 2024 build-out certification guidance among the documents it replaces, and states that it rescinds previous guidance and policies not required by clear and express statutory language, citing the Secretary of Transportation’s memorandum of 29 January 2025.
The substantive change is easy to miss because it reads as a softening of tone. Earlier guidance had made the program’s defining requirement geometric: charging every 50 miles along a state’s portion of the Interstate system, within one travel mile of the Interstate, with exceptions available only in very limited circumstances. The 2025 guidance replaces that with a paragraph headed “Charging Station Distances along AFCs” telling states they should consider the appropriate distance between stations, and may weigh grid capacity, geography, cost and other location-specific constraints. It also streamlines what FHWA reviews in a state plan down to three items and says outright that anything a state includes beyond the legally required information is at its own discretion.
The technical floor did not move with it. The eCFR versioner service reports that every section of 23 CFR part 680, including the four-port and 150 kilowatt provisions and the 97 percent uptime provision, still carries an amendment date of 30 March 2023. So the federal charging network now has a fixed engineering specification and a discretionary map. A station will be built to the same electrical standard everywhere and the decision about where it goes has moved from Washington to the state capitals, which changes who is accountable for a gap in coverage.
Two federal counts, one live database
The most reliable public inventory is the Alternative Fuels Data Center station locator, and the honest way to cite it is with a retrieval date, because it is a database rather than a publication. Retrieved on 13 August 2026, its public-access table gave 81,621 station locations in the United States with 255,073 charging ports, of which 74,349 were direct current fast ports and 180,017 were AC Level 2. On the same day the Joint Office of Energy and Transportation reported 258,800 public ports nationwide, more than double the 2020 figure. The two numbers differ by about 1.5 percent and neither is wrong; they were drawn from a moving dataset at different moments, and that is the level of precision this subject supports.
The federally funded subset is reported separately, and it is small. Through September 2025, the Joint Office’s EV-ChART summary covered stations built under the formula program, the discretionary grant program and the reliability accelerator together, and reported 322,000 charging sessions delivering 12,400,000 kilowatt hours and enabling 37,200,000 electric vehicle miles. Divide the energy by the sessions and the average federally funded session delivered about 38 kilowatt hours, which at 150 kilowatts is roughly fifteen minutes of charging and, on any current highway vehicle, a partial fill. Divide the miles by the energy and the summary is assuming about 0.33 kilowatt hours per mile. Both derived figures matter, because a corridor sized for partial fills is a different building than one sized for full charges.
The bottleneck is an interconnection, and the evidence about it is indirect
The National Renewable Energy Laboratory’s 2023 assessment of 2030 demand put a capital figure on the national network: roughly $82 billion in public and private charging infrastructure for a scenario in which half of light-duty sales are electric and 33 million electric vehicles are on the road, about three times the laboratory’s estimate of investment planned at the time. The sentence that matters for corridor planning is the scope note. Those costs include charging equipment and installation and exclude the cost of grid upgrades and distributed energy resources. The headline number leaves out the part that takes the longest.
How long is genuinely hard to source, and the temptation is to reach for the wrong dataset. Lawrence Berkeley National Laboratory’s Queued Up series is the best-documented account of American interconnection delay: as of the end of 2024, around 10,300 projects were actively seeking grid interconnection, the median duration from interconnection request to commercial operation had doubled from under two years for projects built between 2000 and 2007 to over four years for those built between 2018 and 2024, and of the capacity requested between 2000 and 2019, only 13 percent had reached commercial operation by the end of 2024 while 77 percent had been withdrawn.
Those figures describe generators seeking transmission interconnection. A charging plaza connects as load, through a distribution utility, under a different process, and the Berkeley numbers cannot be transferred to it. What they establish is the state of the institutions on the other side of the meter: the same utilities, the same substations and the same long-lead equipment. For a specific site, the checkable quantities are the distribution utility’s published line extension and service upgrade tariff, its interconnection study queue for large loads, and the lead time its transformer supplier quotes. Those are the numbers a corridor program should be publishing, and no federal source currently aggregates them. That pattern repeats through more on transportation technology, where the decisive quantity is routinely held by somebody under no obligation to publish it.
What a charging desert now means
With the 50-mile rule out of guidance, the phrase has no federal definition left. What remains is a designation, the Alternative Fuel Corridor, and a station standard that applies to anything built along one with these funds. Coverage can still be tested, using the corridor designations alongside the state-level fast charging counts in the station locator, but the test is now against a state’s own stated plan rather than against a national spacing rule.
Interstate fuel and food have always been privately built on publicly designated land, sited where the traffic and the utility made a station pay. Charging built on the same logic is a return to form rather than a departure. The shortfalls that arise from it are recorded in the truck parking shortage, and the smart highway systems share the same roadside cabinets and the same power supply. Electricity adds one condition the fuel model never had: the second network the site depends on has a queue of its own, and no highway agency controls it.