The Tesla Footage Spreading Across America

A Charging Network Is Counted in Ports, Not Locations

Any claim about charging infrastructure rests on a federal dataset, and that dataset counts two different things which are routinely quoted as though they were one.

A station is a location. A port is an individual outlet at which one vehicle can charge. A single station can hold two ports or twenty, so a count of stations and a count of ports produce very different pictures of the same network.

The distinction matters because what a driver experiences is ports, not locations. Waiting behind one vehicle at a two-port site is a different proposition from arriving at a site with sixteen.

There is a second split inside the port count. Level 2 alternating-current ports and direct-current fast ports serve completely different use cases — one is measured in hours, the other in tens of minutes — and pooling them produces a number that describes neither.

A third variable is availability. A port that exists is not a port that works, and reliability studies have repeatedly found a meaningful share of public fast chargers non-functional at any moment, which no inventory count captures.

Networks add a fourth layer. Access, pricing and payment differ by operator, and a port on a network a driver cannot use is not available to that driver in any practical sense.

The federal dataset is public, updated continuously, and carries all of these fields, which means anyone quoting a single headline number has chosen which ones to drop.

None of this is an argument about any vehicle or any company. It is about how a network is counted and what an aerial photograph of a parking lot can establish, which is very little.

Stations, Ports and Levels

The structure of the data determines what any count can mean.

Historical snapshots make the series usable for trends. The dataset is versioned, so counts at earlier dates can be compared with today's on the same definitions, which is the only way to say whether a network grew or whether the counting method changed.

Level 1 charging uses an ordinary household outlet and adds a few miles of range per hour, which makes it relevant for overnight use and irrelevant for travel.

Level 2 operates at higher voltage and typically restores a vehicle over several hours, which suits workplaces, hotels and destinations where a car sits anyway.

Direct-current fast charging bypasses the vehicle's onboard converter and delivers far higher power, and it is the only category that supports long-distance travel on a schedule.

Standardization has been moving toward a common connector across most of the market, with adapters bridging the gap during the transition, and the dataset records which standard each port carries so the mix at any moment is visible rather than assumed.

Connector standards complicate the count further, since not every fast port fits every vehicle without an adapter, and the dataset records connector types for exactly this reason.

Vehicle limits cap the benefit in any case. A car that accepts a maximum of a hundred and fifty kilowatts draws that much regardless of the port's rating, and charging slows sharply as the battery fills, so the headline power of a site describes its best case rather than a typical session.

Power ratings vary widely within the fast category, and a fifty-kilowatt port and a three-hundred-and-fifty-kilowatt port deliver very different experiences while counting identically in a simple total.

Access fields matter as much as hardware. The dataset records whether a location is public or private, whether it is restricted to fleets, residents or customers, and what hours it operates — and a port behind a gate counts in a national total while serving nobody passing through.

What Reliability Studies Show

Installed capacity and working capacity are different quantities, and the gap has been measured.

Independent studies that visited public fast charging sites and attempted charges have reported meaningful failure rates, with problems ranging from payment systems to communication errors to physically damaged cables.

Operators report higher uptime than third-party testing finds, which is largely a definitional difference: uptime measured as the charger reporting itself available is not the same as a driver completing a session.

Sample size is the caveat on all of it. Field studies covering a few hundred sites in one metropolitan area are not a national estimate, and the variation between operators in those studies has been larger than the variation between regions.

Federal funding programs attached to charging deployment have written minimum uptime requirements into their terms precisely because of this gap, with reporting obligations to match.

Redundancy is the design answer to both problems. A site with eight ports tolerates one failure in a way a site with two does not, which is another reason the port count carries more information than the station count.

Queueing matters alongside reliability. Utilization data shows that a small share of sites absorb a large share of sessions, so average availability across a network conceals the congestion at the sites people actually use.

Payment and roaming are the other recurring failure point. A driver without the right application or card can stand in front of a working charger and be unable to start a session, which is why interoperability requirements have been written into public funding terms.

Maintenance economics sit underneath. A fast charger is a piece of industrial power electronics in an unattended outdoor location, and the cost of keeping it working is a recurring expense rather than a one-time build.

What a Photograph Cannot Establish

Images of parking areas circulate constantly as evidence about infrastructure, and they carry almost none.

Equipment installed and equipment energized are different states, and the interval between them can run to months while utility interconnection is completed.

Timing compounds it. Charging demand is concentrated at particular hours and around holidays, so an image taken mid-morning on an ordinary weekday shows the least informative moment of the week.

Occupancy at an instant says nothing about utilization, which is measured in sessions and kilowatt-hours over time and is reported by operators and by federal programs.

Fleet and depot charging is invisible in the same way and is growing faster than public infrastructure in several categories, because a vehicle that returns to the same yard every night needs no public network at all.

Site selection determines what is visible. Charging at home and at workplaces accounts for the majority of energy delivered to light-duty electric vehicles, and none of it appears in any photograph of a retail parking lot.

The comparison being invited — that a parking area shows something missing — is a claim about a distribution, and a distribution cannot be observed from one frame.

What can be observed from data is straightforward: the count of ports by level and network at that coordinate, which is in the public dataset, updated continuously and downloadable.

What Would Actually Show a Network Filling In

Five measurable series describe the build-out, and none of them requires a camera.

Port counts by level, tracked over time from the federal dataset, which publishes historical snapshots as well as current records.

Utilization, reported through federal program requirements and by network operators, measured in sessions and energy delivered rather than in equipment installed.

Interconnection and energization, since a site that has equipment but no utility connection is not yet part of the network.

Price per kilowatt-hour is worth tracking alongside, since it varies by network, by state and by time of day, and it determines whether public charging competes with fuel on the only basis most drivers care about.

Uptime as measured by successful sessions rather than by self-reported availability, which the funding programs now require to be reported.

Grid capacity behind the site is the constraint nobody photographs. A location's ability to add ports depends on the transformer and feeder serving it, and upgrades to those run on the utility's timetable rather than the site owner's.

And power delivered per site, because a location's capacity to serve traffic depends on the total kilowatts available rather than on how many cabinets are standing there.

Geography is the final dimension the single number hides. Coverage along interstate corridors, in dense urban areas without driveways, and in rural counties are three different problems with three different economics, and a national total speaks to none of them.

The composite point is that charging infrastructure is counted in two different units that are routinely conflated, that installed and working are different states with measured gaps between them, and that the dataset answering all of it is free and updated daily.