Drivers report far more trouble with public charging than with the vehicles themselves. The reason is that a charging session depends on a chain of systems, any one of which can fail alone.
The equipment lives in a hostile place
Chargers sit outdoors permanently, exposed to rain, road salt, temperature swings and physical impact from vehicles manoeuvring around them. Everything about that environment shortens component life.
Connectors and cables are handled roughly by many different people, and high-power cables are heavy and stiff, which means they get dropped and dragged across ground repeatedly.
Unlike a vehicle, which has one owner who notices problems immediately, a charger can sit faulty for weeks unless someone reports it or remote monitoring catches the failure.
High power makes components work hard
Fast charging moves very large amounts of power, which requires substantial conversion hardware, cooling systems and often liquid-cooled cables to keep connectors at safe temperatures.
Each of those subsystems is a source of failure, and thermal management in particular tends to degrade gradually rather than break cleanly, producing sessions that start and then slow.
Because the equipment is expensive, sites often carry limited redundancy, so a single failed unit removes a meaningful share of a location's capacity.
Payment and networking are separate failure paths
A session typically requires authorisation through a network connection, and a charger that is electrically healthy will refuse to start if that connection is unavailable.
Card readers, mobile applications and roaming agreements between networks each add a step, and a driver using a network they do not have an account with depends on all of them.
This is why so many reported failures involve a charger that appears functional. The fault is in the transaction rather than in the power electronics.
Communication between car and charger is negotiated
Before power flows, the vehicle and charger exchange messages agreeing on limits and monitoring safety conditions throughout the session.
Implementations differ in small ways across manufacturers and firmware versions, so certain combinations fail while both sides work correctly with everything else.
Ownership is split across several parties
The site host, the network operator, the hardware manufacturer and the electricity supplier are often different organisations with different priorities and response times.
A fault therefore requires coordination before it can be repaired, which lengthens outages compared with equipment maintained by whoever owns the site.
Networks that own their hardware and locations tend to report better reliability for this structural reason rather than because their equipment is inherently different.