What Actually Causes
EV Charger Downtime in India

A driver stands at a charger that says "available" in the app. They plug in. Nothing happens. Was the charger down? Maybe — or maybe the payment gateway timed out, or the backend never cleared a session that failed to close ten minutes earlier, or the earthing resistance drifted high after last night's rain. "Uptime" is usually reported as a single number, but it's actually measuring the weakest link in a chain of eight separate systems, and knowing which one breaks — and why — is the difference between fixing the actual problem and just restarting the charger and hoping.

Downtime is an end-to-end problem, not a hardware problem

A useful way to think about a charging session: grid supply → site electrical system → charger power electronics → cable and connector → vehicle handshake → network connectivity → operator cloud → authorisation and payment. Any one link failing takes down the session, even if every other link — including the charger's power electronics — is working perfectly. A CMS dashboard reporting 99% "online" doesn't mean 99% of sessions succeed; it usually just means the charger has a network heartbeat, which is a different and much easier bar to clear than "a driver can actually plug in and charge."

That distinction matters more in India than most places right now. A widely cited industry snapshot from early 2024 found close to half of public chargers were non-functional at some point in a given month, and roughly a quarter faced frequent downtime tied to technical faults, grid instability, or delayed maintenance — with connector wear, protocol mismatches, payment failures, and unstable connectivity named as the recurring causes. The industry-accepted uptime benchmark sits around 97%, and the gap between a reported dashboard number and what a driver actually experiences is exactly where charging networks lose trust.

The eight places it actually breaks

• Grid supply. India's low-voltage distribution network wasn't built with EV load in mind. Voltage sags, phase imbalance from single-phase residential charging, and harmonic distortion from charger power electronics are all real, documented effects — IEEE 519-2014 sets harmonic limits specifically because of this. Weaker grids in smaller towns and older buildings are the most exposed.
• Site electrical system. Weak earthing and informal connections are a persistent, India-specific safety and reliability risk, especially outside major metros. Earthing resistance is supposed to stay under roughly 10 ohms; if it drifts higher — often after monsoon moisture ingress — a charger's earth-fault protection will trip repeatedly, and repeatedly resetting it without fixing the earthing is a genuine hazard, not a fix.
• Charger power electronics. Thermal derating is a designed-in safety behaviour, not a fault: if internal or cable temperature exceeds a safe limit, the charger reduces power or stops the session rather than risk damage. In Indian summers, ambient temperatures above 45°C make this a routine occurrence for chargers not specifically rated for it — which is why IP and thermal ratings on a spec sheet aren't a formality.
• Cable and connector. Physical wear from repeated plug/unplug cycles is one of the most common hardware-side failure points industry-wide, and it's also one of the most preventable with routine inspection rather than run-to-failure maintenance.
• Vehicle handshake. Not every failed session is the charger's fault. BMS communication quirks, a vehicle-initiated abort, an unseated connector, or old vehicle firmware can all end a session even when the charger itself functioned correctly — which is why fault logs need to separate charger-side failures from vehicle-side ones, or maintenance teams end up chasing problems they can't actually fix.
• Network connectivity. A charger that loses its link to the backend can't authorise a session or process payment, even with a perfect electrical supply. Cellular connectivity in India varies widely by site, and this is a common, underappreciated failure point outside dense urban cores.
• Operator cloud / CMS. This is where a previous session failing to close cleanly can leave a connector showing "occupied" or "faulted" in the app when it's physically free — a software bug presenting as a hardware outage to the driver standing in front of it.
• Authorisation and payment. Payment gateway timeouts, RFID reader faults, or app/backend issues can block a session that has nothing wrong with it electrically. Fragmented software layers — different vendors for hardware, CMS, and payment — are a common source of exactly this kind of failure, since no single team owns the full chain end to end.

Why this framing matters for evaluating a CPO or CMS vendor

Any vendor can quote an uptime percentage. Few can tell you what it's actually measuring, or show a fault log that separates the eight causes above instead of bucketing every failed session as "charger down." The more useful questions to ask a vendor: does your uptime number reflect network-heartbeat availability, or actual successful-session rate? Do you separate vehicle-side failures from charger-side ones in your fault reporting? What's your actual field response SLA, broken down by fault type — not just a single average?

For what it's worth on our own numbers: Massive's CMS-managed network runs at 99.4% uptime, measured at the network level rather than a single-site best case, with an 8-hour average field response SLA on DC faults and 24 hours on AC, backed by engineers across 38 cities. That's not a claim that nothing ever breaks — it's a claim about how fast the chain gets fixed when something in it does.