Uptime used to be a commercial preference for charge point operators. It is now a legal number. The Public Charge Point Regulations 2023 require rapid networks of 50kW and above to be reliable for 99% of the calendar year, averaged across the network, with a report submitted to the Secretary of State every January. That single figure quietly redefines what a charging business is. It is no longer an energy retail operation with some hardware attached — it is a distributed maintenance operation that happens to sell electricity. Hitting 99% is not a dashboard exercise or a reporting one; it comes from knowing the true condition of every unit, catching faults before they become outages, and closing repairs against a clock. That is the operating discipline OxMaint brings to charging networks through asset-level maintenance management.
99% Is Not a Target. It Is a Maintenance Schedule.
OxMaint turns fault codes, inspections and SLA clocks into tracked work orders against every chargepoint, so availability is engineered rather than hoped for.
What 99% Actually Buys You in Hours
The most useful thing an operations team can do with the reliability requirement is convert it out of percentages and into hours, because percentages hide how tight the budget really is. A year contains 8,760 hours. One percent of that is 87.6 hours. That is the entire annual downtime allowance for an average unit across the network — including planned servicing, firmware work, vandalism, cable replacement, payment terminal failures, grid outages and the time a technician spends driving to site.
Put another way, a single unit that sits dead for one long weekend has consumed roughly three-quarters of its yearly allowance. Two such events and it is over budget, carried only by the rest of the network averaging it down. On a small network of twenty rapid units, there is very little averaging available. On a network of four hundred, one chronically broken site can be absorbed — which is exactly how operators end up with a compliant headline number and a handful of locations that drivers have learned to avoid.
Framed as 7.3 hours a month, the number changes how you plan. A four-hour planned service visit is not a routine event — it is more than half of that month's budget. Suddenly it matters a great deal whether the technician arrives with the right part, whether the fault was correctly diagnosed remotely before dispatch, and whether the visit closes the issue or generates a return trip. These are maintenance management problems, and they are solvable.
A Chargepoint Is Seven Systems Pretending to Be One Machine
Drivers experience a charger as a single object that either works or does not. Operationally it is a stack of independent subsystems, each with its own failure modes, its own diagnostic signals and its own repair skillset. A fully healthy power module behind a dead card reader produces exactly the same outcome as a blown power module: a driver who leaves.
Understanding downtime means attributing it to the correct layer, because the layers behave completely differently. Some fail gradually and warn you. Some fail instantly with no signature. Some can be recovered remotely and some absolutely should not be. Treating every outage as a generic fault is how networks end up resetting a unit four times a week instead of replacing a contactor once.
The Four Failure Modes That Consume Most of the Budget
Downtime is not evenly distributed across those seven layers. A small number of failure modes account for the majority of lost hours in most networks, and each of them announces itself before it becomes an outage — if someone is watching for the signature rather than waiting for the fault code.
Connector And Cable Wear
Repeated plug cycles, weather and rough handling degrade pins and sheathing. The early signature is intermittent session failures on one connector while its twin behaves perfectly.
Payment Terminal Failure
Card readers and screens fail on their own schedule and are frequently reported by drivers rather than detected by the backend, so the outage clock starts late.
Thermal Derating
Blocked filters, tired coolant and failing pumps reduce delivered power long before a hard fault appears. The unit looks available while delivering a fraction of its rating.
Connectivity Dropout
Loss of the backend link prevents authorisation on hardware that is otherwise fully serviceable, and is disproportionately common at rural and underground sites.
There is a common and expensive habit worth naming here. Remote reset is a legitimate tool for recoverable software and communication faults, but using it as the default response to every fault masks physical problems and adds stress to contactors and power electronics. The dashboard looks healthier for a few hours while the underlying asset degrades faster. A mature network separates what can be reset from what must be diagnosed.
Stop Resetting the Same Charger Every Tuesday
OxMaint tracks fault history per unit, so repeat offenders surface as a pattern and get a root-cause repair instead of another remote reboot.
The Restoration Clock: Where Hours Are Really Lost
Most operators measure response time. Very few measure the full clock, and the gap between the two is where the annual allowance disappears. A fault detected at 09:00 and attended at 13:00 looks like a four-hour response. If the technician arrives without the correct part, orders it, and returns eight days later, the unit was down for over a week. The response SLA was met. The reliability figure was destroyed.
| Clock Segment | What It Measures | Typical Loss | How It Gets Shortened |
|---|---|---|---|
| Detection | Fault occurring to fault known | Hours, if reliant on driver reports | Automated fault ingestion and heartbeat monitoring |
| Triage | Fault known to correctly classified | Hours, when every fault looks alike | Severity rules mapped to fault class and site priority |
| Dispatch | Classified to engineer assigned | Hours to a full day | Skill and geography-based assignment from the work queue |
| Travel | Assigned to on site | Unavoidable, but routable | Batching nearby jobs and prioritising by network impact |
| Diagnosis | On site to cause identified | Hours, without history | Full asset and fault history available on the mobile job |
| Parts | Cause identified to part in hand | Days to weeks | Spares linked to asset models with stock visibility |
| Repair and return | Part in hand to unit available | Hours, plus any second visit | First-time-fix tracking and procedure-led repairs |
Read down the "typical loss" column and the conclusion is uncomfortable but useful: the biggest single loss is almost never the repair itself. It is parts. A network that has not linked spares to specific charger models is effectively buying downtime every time a power module or a connector assembly fails, and no amount of engineer responsiveness compensates for it.
Planned Servicing That Fits Inside the Allowance
Because planned work counts against availability in most reliability calculations, preventive maintenance on a charging network has to be designed around minimum intervention time rather than maximum thoroughness. The aim is short, frequent, well-prepared visits rather than long annual overhauls.
Visual And Housekeeping
Cable and connector condition, screen legibility, signage, bollard damage, lighting, litter and drainage. Fast, low-skill, often combinable with site host activity.
Functional Verification
Test transaction through the payment path, connector pin inspection, locking mechanism check, emergency stop verification and reset, delivered power spot check.
Thermal And Enclosure
Filter cleaning or replacement, fan operation, coolant level and condition on liquid-cooled units, seal and gasket integrity, internal cleanliness.
Electrical And Certification
Protective device testing, earthing and insulation verification, torque checks on terminations, statutory electrical inspection and documented certification.
The scheduling detail that matters most is timing. Servicing a motorway rapid site at 11:00 on a Friday costs far more availability than the same work at 04:00 on a Tuesday, even though both consume identical engineer hours. Networks that schedule preventive work against site-level demand profiles rather than against a flat calendar reclaim a meaningful slice of their annual budget for nothing.
We were hitting our response SLA on more than ninety per cent of faults and still sitting well below where we needed to be on availability. The problem was never the engineers. It was that a third of visits ended in a parts order, and each of those turned a four-hour outage into a two-week one. Linking spares to charger models did more for our uptime figure than any change to the rota.
How OxMaint Supports Charging Network Operations
OxMaint treats every chargepoint as an asset with a service history, a maintenance schedule, a spares list and a performance record. That structure is what lets fault data, planned work and repair evidence live in one place instead of being split between a monitoring platform, a field service app and a spreadsheet of SLA breaches.
Chargepoint Asset Register
Every unit and connector recorded by serial, model, site and commissioning date, giving each one an individual availability and fault history rather than a network average.
Fault To Work Order Automation
Incoming faults raise classified work orders by severity and site priority, so the triage decision is made by a rule rather than by whoever happens to be watching the queue.
SLA Clocks And Escalation
Response and restoration targets run per job with automatic escalation, and the full clock is recorded segment by segment rather than collapsed into a single closure timestamp.
Preventive Schedules By Model
Weekly, monthly, quarterly and annual regimes generate automatically per charger type, with procedures, safety steps and required parts attached to each task.
Spares Linked To Models
Connectors, modules, filters and payment components associate with the units that use them, so stock decisions are driven by installed base rather than guesswork.
Availability Reporting
Downtime tracked per unit and rolled up across the network, with planned and unplanned hours separated, ready for annual compliance reporting and internal review.
The Four Numbers Worth Watching Monthly
Network availability is a lagging indicator. By the time it moves, the cause is months old. These four leading measures move first and tell you whether next year's number is being built or eroded.
First-Time Fix Rate
The share of faults resolved on the initial visit. Every failure here multiplies a short outage into a long one and is usually a parts or diagnosis problem.
Repeat Fault Rate
Units generating the same fault class more than twice a quarter. A high figure means resets are substituting for root-cause repair somewhere in the estate.
Mean Time To Restore
The full clock from fault detected to unit available, not from dispatch to attendance. This is the number that actually converts into lost hours.
Worst-Ten Concentration
How much of total downtime comes from your ten worst units. If it is most of it, the network problem is really a small hardware problem.
Frequently Asked Questions
What uptime are rapid charge point operators required to achieve?
Rapid networks of 50kW and above must be reliable for 99% of each calendar year, measured as an average across the operator's rapid network, with performance published and reported annually.
Does planned maintenance count against reliability?
A unit that is not working counts as unavailable regardless of why, so planned work consumes the same allowance as a fault. This is why short, well-prepared, off-peak visits matter so much. OxMaint schedules preventive work around demand profiles.
Why does my network average look fine while drivers complain?
Because a network average hides concentration. A handful of chronically faulty units can be absorbed statistically while ruining the experience at specific sites. Tracking availability per unit exposes this immediately.
Is remote reset an acceptable maintenance strategy?
It is a valid tool for recoverable communication and software faults only. Using it for physical, thermal or insulation faults hides the real issue and adds wear to power electronics over time.
What single change improves uptime fastest?
Linking spares to charger models so engineers arrive with the right part. Second-visit rates are usually the largest recoverable source of lost hours. Book a demo to see spares mapped to assets.
Make Next January's Reliability Report a Formality
Track every chargepoint as an asset, run preventive work that fits the allowance, close faults against a real restoration clock, and watch availability become something you manage rather than measure.
