Solar Inverter Failure: Safe First Actions

A stopped inverter feels like an emergency, and the first instinct, pressing reset again and again, is usually the wrong one. This guide gives the order of work that protects people first, evidence second and generation third: what to check from the ground, how to tell a grid problem from an equipment problem, what to record, who to call, and what a downtime day actually costs. It also covers the quieter cases where an inverter keeps running but loses output, because those cost as much as a clean stop.

Responding to a fault cannot be reduced to a brand score or a universal recovery deadline. Reliability depends on the exact model, the installation, the grid, the environment, maintenance, firmware, and how a “failure” is defined and counted. Comparable, independently audited field data is not available for the brands discussed on this site, so this guide gives a method, not a statistic.

The first 15 minutes: safe actions in order

Do these from the ground, in this order, and stop at the first step that finds the cause.

1. Look and smell before you touch. Burning smell, smoke, scorch marks, water on or around the inverter, damaged cable insulation or a swollen enclosure mean stop here. Keep people away, switch off at the AC breaker only if you can reach it without going near the damaged equipment, and call a qualified professional.
2. Check whether the building has mains supply. An on-grid inverter must disconnect when the grid fails, so a power cut stops it by design. This is the most common reason an inverter “is not working”. If the house has no supply, wait for the DISCOM to restore it. The inverter will reconnect after its start-up checks, which take some time.
3. Read the display and LEDs. Photograph the full message and the LED pattern before changing anything. Do not repeat a reset until you have recorded it, because clearing the screen can erase the first piece of evidence a service engineer needs.
4. Check the switches you own. Confirm that the AC breaker and, where you have access, the DC isolator are in the positions your installer left them. A tripped breaker or an isolator turned off during cleaning looks like a dead inverter. Do not close a breaker that has tripped more than once without finding out why.
5. Check the time and the light. An inverter produces nothing at night and little before sunrise or under heavy cloud. A low-light message at dawn is not a fault.
6. Open the monitoring app. Compare its status with the local display. An app can show “offline” while the inverter generates normally, which is a communication problem, not a failure. The Wi-Fi troubleshooting guide covers that case.
7. Note what changed recently. Storms, electrical work, a new appliance, a grid outage, cleaning, or a heat wave. Recent events narrow the cause quickly.

What not to do

• Do not open the inverter or a DC enclosure, and do not unplug rooftop PV connectors. Switching off the AC side does not make illuminated PV conductors safe.
• Do not reset isolation, residual current or arc fault messages. They point to an insulation problem in a circuit that is energised whenever the sun shines. See the error codes guide.
• Do not raise grid voltage or frequency limits to stop trips. Those settings belong to the approved configuration under the utility’s conditions.
• Do not bypass a tripped protective device, and do not send an untrained person onto the roof.
• Do not wait for the next sunny day to see “if it fixes itself” when there is a safety indicator.

Separate external conditions from equipment faults: the diagnosis ladder

A stopped inverter may be responding to a grid outage, grid voltage or frequency outside its permitted window, an open isolator, a protection-device trip, insufficient irradiance, a communication failure, a DC-side fault, overheating or an internal condition. Work down the ladder. Each rung rules out one family of causes.

If the ladder ends at a message you cannot place, the troubleshooting guide lists 15 common problems with the safe check for each, and the error codes guide explains how to group a fault code when you do not have the manual. A unit that repeatedly stops and restarts has its own page: solar inverter keeps switching off.

Five kinds of downtime, and why they are not equal

“Downtime” covers more than a stopped inverter. Naming the kind helps you decide how urgent it is.

• Hard downtime. The inverter has stopped and needs service. Generation is zero.
• Soft downtime. The inverter runs but at reduced output, for example thermal derating or an MPPT input that has dropped out. It is easy to miss because the display looks normal.
• Nuisance downtime. Trip-and-restart cycles from grid disturbances. Each trip is the protection working, but repeated trips cost generation and can point to a weak feeder or a long AC cable.
• Scheduled downtime. Planned maintenance. It is a cost you choose, and it prevents the other kinds.
• Hidden downtime. A gradual loss of output that nobody notices without comparing against earlier months or a clear-day baseline.

Hidden and soft downtime are why a monitoring app matters. The monitoring app guide explains what to watch.

Why inverters stop: five cause groups

Causes fall into five groups. None of them has a universal frequency, because the mix depends on the product and the site.

1. Component ageing. Heat ages capacitors and fans, and thermal cycling stresses power semiconductors. This is a long-horizon cause. The circuit diagram guide shows where each part sits.
2. Grid conditions. Voltage and frequency excursions, switching transients and lightning-induced surges. Many apparent “inverter failures” are the inverter correctly disconnecting from a poor supply, and anti-islanding is the reason an on-grid unit stops during a power cut.
3. Environment. Dust that blocks cooling, moisture entering through damaged seals or cable entries, salt air near the coast, and high ambient temperature. Chronic overheating is covered in the inverter overheating guide.
4. Installation quality. Loose terminations that heat up over time, poor ventilation, undersized cables with high voltage drop, weak earthing that reduces surge protection, and string design that suits summer but not a cold morning.
5. Missed maintenance. Skipped inspections, ignored monitoring alerts and deferred cleaning let small issues grow.

What a downtime day costs: the arithmetic

You can estimate your own loss without relying on anyone’s averages. The loss for a day is the energy the system would have produced multiplied by the value of that energy to you.

The practical point is the detection time. A fault discovered on the day costs a day. A fault discovered at the next bill can cost weeks, which is why alerts and a habit of glancing at the app matter more than any single hardware choice.

What to record before you call anyone

A good fault report shortens every later step. Record these before changing the system:

1. Model name and serial number from the label.
2. Exact message and LED pattern, photographed.
3. Time of the fault and what generation looked like in the hours before.
4. Grid status: was there a power cut, voltage problem or storm?
5. Recent work on the system: cleaning, electrical work, new loads.
6. Anything unusual: sound, smell, heat or visible damage.
7. The array layout if you know it: modules per string and number of strings.
8. Your location and contact details.

Manufacturers ask for much the same list. A Growatt manual, for example, asks for serial number, model, error message, grid voltage, DC input voltage, whether the failure can be reproduced, whether it happened before and the ambient conditions. A Solis manual asks for serial number, dealer, installation date, a description including LED status, the array configuration and contact details. Having it ready turns a first call into a diagnosis.

Safe escalation path

1. Installer or EPC first, if you have one. They know the as-built design, and many faults are string or wiring problems that need their drawings.
2. The manufacturer’s support channel. For Qbits equipment, raise a request on the Qbits support helpdesk{target=“_blank” rel=“noopener”} with the model, serial number, location and the fault as displayed. Ask for the fault code manual for your model if the display message is unclear. The datasheet library confirms which product family you have, and the on-grid inverter and hybrid inverter pages describe the families. A hybrid unit also reports battery and backup-output messages that an on-grid unit never shows.
3. A service partner near you. The authorized service partners directory lists partners by state and district. Confirm scope, coverage and any charges with the partner before work starts, because listings do not guarantee availability or response time.
4. The DISCOM, for grid problems. Repeated over-voltage or outages belong to the utility. Send them your timestamps and displayed voltage readings.
5. Get a ticket number and a written diagnosis. Use the ticket number in every follow-up. Ask what was found, what was done, and what warranty clause, if any, applies.

If there is a burning smell, smoke, water ingress, damaged cabling, repeated protective-device operation or a suspected insulation or arc fault, keep people away and have a qualified professional make the site safe before anything else.

What the service ticket should show

Qbits does not publish a universal service deadline, so ask for the timing in writing for your case and keep it with the ticket.

Repair or replacement: what decides it

The service provider decides after diagnosis, and the answer depends on the model, the fault and the written warranty for your unit. The factors that usually matter:

• The fault. A fan, connector or wiring defect is often repaired on site. Damage to power electronics from water ingress, a surge or burning may point to replacing the unit.
• Repeat faults. A unit with several different faults in a short period may be treated differently from a single isolated one.
• Cause outside the inverter. If the real cause is a string design error, a wet junction box or a poor earth, replacing the inverter will not fix it. Ask for the root cause in writing.
• Warranty terms. The public datasheets describe an expandable warranty but do not define the base term, the remedy, a registration deadline or exclusions. Obtain the current written warranty for your exact model, and read the warranty claim guide before agreeing to chargeable work.

Photograph any physical damage before a technician arrives.

Reliability evidence worth requesting

When someone quotes a failure rate, a “most reliable” label or a service promise, ask what sits behind it.

Without this context, labels such as “most reliable”, “lowest RMA” or “fastest service” are marketing claims, not comparable evidence.

Preventing the next stop

Most prevention is routine and inexpensive. A reasonable pattern, to be adapted to your manual and your installer’s advice:

The inverter maintenance guide, the annual maintenance checklist and the pre-monsoon inspection checklist give the detail. For anything that involves opening the unit or touching DC wiring, use a qualified technician and the manufacturer’s service channel.