• California Energy Commission field tests found 96.2% of Level 2 charging ports and 96.4% of fast-charging ports working when testers arrived.
  • Paying was the weak link: credit and debit cards failed on 34.4% of first payment attempts, while plug-in payment failed on 2.3%.
  • The statewide average leans heavily on Tesla, whose fast chargers worked 99.6% of the time, against 92.1% for all other networks combined.

The statewide figures are a clear improvement on the last round of independent testing, which put Bay Area fast-charger success as low as 72.5%. The gain is uneven: one network accounts for most of it, and the card reader is now the most likely point of failure at the station.

The numbers come from a California Energy Commission staff report published in November 2026, based on University of California, Davis field tests from January 2024 to February 2026. Testers plugged a car in at 3,571 ports, 1,817 Level 2 and 1,754 fast, and inspected 12,586 more, out of a state total of more than 216,000 public and limited-access ports through 2025. The operational rate came in at 96.2% for Level 2 ports and 96.4% for fast-charging ports, and the charge success rate, which asks whether a car charged at a port that was working, came in at 93.1% for Level 2 and 94.0% for fast chargers. Assembly Bill 2061, a 2022 California law, ordered the commission to assess charging uptime from 1 January 2025 and every two years after. The commission does not yet have the continuous data its standards require, so the field tests are an interim assessment. What the testers measured is whether a port worked on the day of the visit, rather than how often it is down over a year.

A fast EV charging station at Mundrabilla Roadhouse in Australia
Calistemon / Wikimedia Commons, CC BY-SA 4.0

California EV charger reliability, measured port by port

Earlier studies put fast-charger success at 77% to 86%, and many combined broken ports and failed sessions into one score. The report’s conclusion stays hedged: although methodologies differ, the results suggest the reliability of California’s publicly accessible charging network may have measurably improved in recent years.

Location made little difference: gaps between disadvantaged or low-income communities and the rest ran 0.3 to 1.2 percentage points, and between rural and urban areas 0.1 to 0.8 points, within the margin of error at a 99% confidence level. The averages leave out the ports nobody could reach. Testers could not get to 1,575 Level 2 and 615 fast-charging ports, and nearly 9% of Level 2 and 7% of fast-charging ports were obstructed, by construction, dumpsters or parked gasoline cars.

Measure Level 2 DC fast charging What it means
Operational rate 96.2% 96.4% Share of ports working at the time of testing
Successful charge attempt rate 93.1% 94.0% Share of plug-in tests where the car charged
Plug-in tests 1,817 ports 1,754 ports Part of 3,571 plug-in tests, January 2024 to February 2026

Payment failed far more often than charging

Paying was the weak link. Card readers were the least reliable way to pay. The figures cover the first method each tester tried: across 3,525 first attempts, 17.7% failed. Credit and debit cards failed 380 times in 1,104 tries, a 34.4% failure rate. Swiping a card failed more often than it worked. Plug-in payment, where charging is authorised when the car is connected, failed 19 times in 817 attempts, a 2.3% rate.

In most cases testers eventually got another method to work. The report notes that drivers may not have access to all the payment types available to UC Davis testers, and it lists weather-related failures, poor software integration and terminals on inferior cellular connections as possible explanations. It also cautions against reading too much into plug-in payment, since most of those tests involved one large network using its own vehicles.

Payment method First-attempt failure rate Attempts
Credit or debit card, all uses 34.4% 1,104
Card, by method Swipe 59.2%, chip 52.1%, tap 26.7% Part of the 1,104 card attempts
App, membership card, QR code and other 14.0% Part of 3,525 first attempts
Plug-in payment 2.3% 817

A network can report a high operational rate while a driver still fails to charge, because the port works and the card reader does not. For a driver choosing between networks, the payment failure rate is the more useful number. Home charging avoids the problem entirely, and our guide to Level 2 home charger installation costs sets out what a driveway charger costs. For drivers without a home charger, our guide to owning an EV in an apartment covers the alternatives.

One network lifts the statewide average

Tesla’s fast chargers were operational at 99.6% of the 4,655 Tesla ports testers reached, about 60% of the 7,705 fast-charging ports in the operational count, with charge success of 98.3% across 867 operational Tesla ports. The report leaves the networks unnamed: it credits the largest fast-charging network with those same two rates, against averages of 92.1% and 89.7% for all other networks, and says the leaders in both charger classes raised the statewide averages. The names come from the spreadsheet published with the report.

The other named fast-charging networks sit well below Tesla. Per the spreadsheet, Electrify America came in at 93.5% operational across 602 ports and 94.2% charge success across 171, EVgo at 92.7% across 898 and 88.3% across 266, and ChargePoint’s fast chargers at 90.6% across 650 and 92.6% across 190. On Level 2 the leader is ChargePoint, at 97.0% operational across 5,521 ports and 96.5% charge success, against 94.9% and 85.6% for all other networks. Even without the largest network, the report says, reliability exceeded what past studies reported, though the variance shows clear room for improvement at lower-performing networks.

Network Operational rate Charge success rate Ports in operational count
Tesla, DC fast charging 99.6% 98.3% 4,655
Electrify America 93.5% 94.2% 602
EVgo 92.7% 88.3% 898
ChargePoint, DC fast charging 90.6% 92.6% 650
ChargePoint, Level 2 97.0% 96.5% 5,521

The spread between networks is the finding that matters most for drivers. A 7.5 percentage point gap separates Tesla’s 99.6% from the 92.1% average of every other fast-charging network, and the gap widens to 8.6 points on charge success. Drivers do not experience a statewide average. They experience one network at one station. Connector type adds another variable: our explainer on NACS versus CCS charging connectors sets out which plug fits which car, and our analysis of 22,700 cars and their battery life covers how much range a driver needs from a public charger.

What the 96.4% figure says about the 97% rule, and what it does not

The commission’s own performance standard does not use either headline number. Under regulations the Office of Administrative Law approved in part on 10 March 2026, effective 1 April, publicly or ratepayer funded fast chargers installed on or after 1 January 2024 must maintain at least 97% annual average uptime for each calendar year of their first six years. Level 2 chargers are exempt, along with fleet, private residential and several other classes. The bar counts minutes of downtime across a year. The field test checked ports on a single visit, and the report says it does not determine uptime, so 96.4% is a snapshot rather than a near miss.

A network could score 96.4% on visit days and still clear 97% annual uptime if its outages are short, or fail the standard badly if its outages are long. The field tests cannot tell the difference. Staff will begin collecting the time-series data the regulations require in late 2026, and those data will be the basis of the first official uptime assessment under AB 2061. Until then, the honest reading of the report is narrower than the headline suggests: California’s chargers usually work when someone checks, the payment terminal is the most likely point of failure, and the network you use matters more than the state you are in.

Sources