Backup Power After an Outage: 4 Situations, and the Right Fix for Each

There's no single answer to "my power backup failed"

I coordinate emergency power and equipment deliveries for a mid-size B2B distributor. In the past 14 months, I've handled 60+ rush jobs where something went wrong after a blackout—not before. That's the call nobody plans for.

Here's what those calls have in common: the person on the other end usually thinks they're describing one problem. They aren't. There are at least four different situations that get lumped under "backup power failed," and the fix for each one is completely different.

Diagnosing the actual situation first—instead of guessing and buying parts—is the single biggest efficiency gain in this whole process. I'll walk through the four scenarios, then help you figure out which one you're in.

Situation 1: Your CyberPower UPS doesn't turn on after a power outage

This is the most common emergency call I get, and also the most misdiagnosed.

Someone's CyberPower UPS—or any standby or line-interactive unit—takes a hit during an outage, and afterward the panel is dead. No lights, no beep, nothing. The instinct is to assume the unit is fried and order a replacement. Nine times out of ten, that's wrong.

Work through these four steps in order before you replace anything:

1. Reset with the load disconnected. Unplug every device, hold the power button for 10–15 seconds, then plug the UPS into the wall and try again. This clears a latched fault state that most models enter after a surge event.
2. Check the battery. Sealed lead-acid batteries in UPS units degrade on a 3–5 year cycle depending on heat and discharge frequency. A swollen or aged battery will keep the whole unit dark. If you've had the unit for more than four years and it's lived in an uncooled closet, this is your likely culprit.
3. Check the input breaker. Most rackmount and tower units have a resettable breaker on the rear panel. Push it in firmly. If it pops back out immediately, you have a real fault—stop and call support.
4. Run the self-test. If the unit has a self-test button or LCD menu, run it. A passing self-test with a dead output usually points to the battery, not the board.

If none of that works, then you're in replacement territory. But in my experience, more than half the "dead UPS" calls I've logged were resolved by steps 1 and 2.

Looking back, I should have asked about battery age on every single one of these calls from day one. It would have cut our average resolution time roughly in half during the first year.

Situation 2: You need a UPS that actually keeps your equipment running

Different problem entirely. This is the person who didn't have a UPS—or had one too small to matter—and now wants to size a proper unit.

If you're in this situation, the decision tree is simpler than most people make it:

Match the technology to the load

Modern PFC (power factor corrected) power supplies in servers and workstations don't play well with simulated sine wave output. They can shut down, overheat, or refuse to accept the transfer. If you're protecting anything with an active PFC supply—which is most enterprise gear made after about 2015—you need a pure sinewave UPS. CyberPower's PFC-compatible line is built for exactly this, and their capacity range spans 375VA up to 3000VA.

A concrete example: the CP1500EPFCLCD

The CyberPower CP1500EPFCLCD is the unit I get asked about most often for small server rooms and network closets. Quick specs worth knowing: 1500VA / 900W capacity, pure sine wave output, PFC-compatible, and it can be mounted as either a rackmount or tower unit. That last point matters more than people think. A unit you can reorient to fit your rack saves you from buying a separate rack shelf, and it means the same SKU works whether you're deploying it in a closet or a proper enclosure.

Size by wattage, not VA

Add up the actual wattage of what you need to keep alive—not the VA rating on the label. A typical small server plus switch plus NAS runs $200–$350W. At that load, a 900W unit gives you enough headroom to ride out short outages and enough runtime to shut things down cleanly on longer ones. Do not run a UPS at 90% load continuously; it kills battery life.

Situation 3: You're looking for a standalone solar generator

Honestly, I don't sell these. When people ask me what the best solar generator to buy is, I tell them upfront that I don't have inventory in this category—but I do have opinions from watching clients buy the wrong one.

There is no "best" solar generator. There's only the right one for your load and your charging situation. Three things to nail down before you buy:

• Actual vs. peak output. A unit marketed as "2000W" very often means 2000W peak for a handful of seconds, with continuous output closer to 1000W. If the continuous rating isn't clearly listed, walk away from that listing.
• Battery chemistry and cycle life. LiFePO4 (lithium iron phosphate) is the current standard for serious portable units. It costs more upfront, but the cycle life is typically 3–5x that of older lithium-ion packs. If you're using this 100+ times a year, chemistry matters more than capacity.
• Charge method practicality. Solar charging sounds great until you factor in panel wattage and daylight hours. A 400W panel in real-world conditions—shade, angle, clouds—might deliver 150–250W of actual input. Size your panel array for half the manufacturer's rating, not the sticker number.

If you're in this situation and you only need 1–2 days of backup for a modest load, you may find that a properly sized UPS plus a small inverter generator is cheaper and more flexible than a large solar generator. That's a judgment call, not a rule—it depends on how often you'd actually use the solar capability.

Situation 4: Your small engine backup won't start when you need it

This one catches people off guard. You've got a generator or a piece of small-engine equipment sitting in the garage as your "backup plan," the power goes out, and it won't turn over. That's not a UPS problem. It's an engine maintenance problem, and it's the most preventable one on this list.

The spark plug is the first place to look

Most small engines that fail to start after sitting for months have a fouled or corroded plug. If your equipment calls for a QL78YC spark plug—a common spec for small engines on trimmers, chainsaws, and some generator models—check it before you touch the fuel system. Pull the plug, look at the electrode. Black and wet means fouling; white or ashy means it's running lean. Either way, a fresh plug is $5–$10 and eliminates a variable fast.

Get the boot off without destroying it

The frustrating part is usually getting the spark plug boot (the rubber connector over the plug) off cleanly—especially on older or heat-cycled engines where it's basically welded on. A spark plug boot puller pliers is the right tool for the job. It grips the boot at the base and pulls straight out without twisting. Using regular pliers, or worse, yanking by the wire, is how people tear the boot or crack the plug ceramic. I've seen more engine-down time caused by a torn boot than by a bad plug.

Store a spare QL78YC and a boot puller with the equipment. It's a $25 insurance policy against a 5-day outage becoming a 7-day one.

How to tell which situation you're in

Here's the quick version. Ask yourself three questions, in order:

1. Was there already a UPS in place, and is it now dead? → Situation 1. Start with reset and battery, not replacement.
2. Was there no UPS, and the outage took down your equipment? → Situation 2. Size by continuous wattage, and check whether your power supply is PFC.
3. Is the failure in a portable power station or a small engine, rather than a plugged-in unit? → Situation 3 (portable/solar) or Situation 4 (engine). If the unit charges from panels, it's 3. If it burns fuel, it's 4.

One caveat, because I'd rather be honest than sound authoritative: my experience is based on roughly 200 mid-size B2B power and equipment orders. If you're running data-center-scale infrastructure or industrial-grade generator systems, the diagnosis logic holds but the specifics may differ. I can't speak to enterprise-tier sizing from this sample.

The bigger point—and where I've changed my thinking over the years—is that the cheapest emergency fix is almost never a replacement part. It's a correct diagnosis. Systematically work through the situation you're actually in, and you'll spend less time and less money than the person who buys first and troubleshoots second.