AC Compressor Failure? 6 Steps to Get Your HVAC Back Online Fast

If you've ever had an AC compressor die on a Friday afternoon, you know that special kind of panic. Facilities managers, maintenance leads, plant engineers—we've all been there. In my role coordinating emergency repairs for commercial buildings, I've handled 200+ rush jobs over 12 years, including same-day turnarounds for hospitals and data centers. A few of those jobs were big enough that missing the deadline would have triggered a $50,000 penalty clause.

Here's the thing I've learned: emergency HVAC situations don't leave room for creative thinking. You need a predictable order of operations. This checklist is what I run through when a client calls at 7:30 AM saying the compressor won't start, or the radiators are cold, or the control panel is throwing faults they can't decode. Six steps, straight through. If you're in that situation right now, start here.

Step 1: Confirm It's Actually the Compressor

The first mistake people make is assuming the AC compressor is dead before running basic diagnostics. In March 2024, a client was sure their 10-ton compressor had seized. Turned out it was a contactor with burned contacts—a $35 part, not a $4,000 compressor. We had that system back online in two hours.

Here's what to check before you order any replacement:

• Power supply. Is the breaker tripped? Are all phases present? Check the contactor with a multimeter.
• Capacitor. A swollen or leaking start capacitor stops a compressor dead. This is the most common failure I see.
• Control signal. Is the controller actually calling for cooling? Compare the setpoint against real room temperature from a reliable sensor.
• Compressor windings. Check resistance across the terminals. A short to ground means it's done. Within spec? Probably not the compressor.

Compressors are workhorses, and a well-built one should run 15+ years. That's true whether we're talking about a building's AC or the tiny unit inside an Omron compressor nebulizer NE-C801 used in clinics worldwide—when compressor engineering is right, it just runs. When it's wrong, it fails at the worst possible moment. Use this step to collect data, not guesses. Write down error codes, take readings, and photograph the wiring before you touch anything.

Step 2: Stabilize the Space While the System Is Down

Once you confirm the system is down, your next job is protecting whatever that space is supposed to protect. If it's winter and you're waiting on a heat pump or boiler part, that means temporary heat.

A buddy heater—the propane radiant type—can be a lifesaver in a small mechanical room or pump house. I've used them to keep pipes from freezing during overnight repairs. But there are rules:

• Never leave it running unattended.
• Keep it at least 3 feet from anything flammable.
• Ventilation is non-negotiable. Propane heaters produce carbon monoxide.
• Put a CO detector in the same space. Always.

If the failure is in cooling (server rooms, process areas, clinic storage), bring in temporary AC units and monitor the temperature closely. You're not trying to make the space perfect—you're buying time until the main system is repaired.

Step 3: Bleed the Radiators (Yes, Really)

Here's a step that gets skipped way too often. If your heating system uses radiators and the heat isn't reaching them evenly—or if you've just replaced a pump or opened any part of the hydronic loop—air is almost certainly trapped inside. Air locks create cold spots and make a perfectly good boiler look broken.

Bleeding a radiator is simple, and it might save you a needless emergency service call. Here's how to bleed a radiator properly:

1. Turn the heating system off and let it cool down. Bleeding a hot radiator is how people get scalded.
2. Locate the bleed valve at the top of the radiator, on one side. It's usually a small square fitting.
3. Open it slowly with a radiator key or a flat screwdriver. You'll hear a hiss as air escapes.
4. Close it the moment water flows steadily. The hiss turning to a dribble means all the air is out.
5. Check the system pressure on the boiler gauge afterward. You may need to top up the loop.

An old misconception: "you should bleed radiators all the time." That's a holdover from an era of poorly sealed cast-iron systems. Modern hydronics only needs bleeding after work that opens the loop. I've seen a maintenance team spend two days chasing a "boiler fault" that was just air locked in one wing. Bleeding is free. Do it before you authorize a pricey emergency repair.

Step 4: Source Genuine Parts Through an Authorized Distributor

If the failure needs a real replacement—say, an Omron power supply or controller—don't get clever with a sketchy listing that's 60% cheaper than everyone else. A counterfeit part might work for a week, then take out the PLC it's feeding. In emergency mode, that turns one day of downtime into three.

Through an authorized Omron power supply distributor, you get three things the gray market can't guarantee: a genuine product, a warranty, and specifications that actually match your system.

Now the money question. When a distributor quotes $120 for next-day delivery instead of $20 for ground, that's not a rip-off—it's a certainty service. I have mixed feelings about rush shipping fees. Part of me thinks they're a tax on poor planning. But another part remembers the project where an $800 overnight freight charge saved a $12,000 contract. And I do not mean every rush fee is justified. But when a deadline is real, speed isn't the luxury. Certainty is.

Step 5: Replace, Reinstall, and Test the Full Loop

New part in hand. Now install it the right way—including testing, where most post-repair failures actually happen.

1. Lock out the power at the disconnect. OSHA's lockout/tagout standard (29 CFR 1910.147) isn't optional, even on a rushed timeline.
2. Install the part and verify every connection against the schematic. Photograph your work before closing panels.
3. Power up and observe startup. Watch the compressor's current draw. An Omron PLC or temperature controller will show you exactly what it's reading from the sensor.
4. Run a full cycle. Let the system pass through the setpoint and verify it cycles correctly. This catches 90% of installation errors.
5. Monitor for 24 hours. A compressor that short-cycles or trips on thermal overload may still have a refrigerant or ventilation issue. Check it the next morning.

Skip the test cycle and you're gambling. I've seen systems run beautifully for 20 minutes, then fail at 2 AM. Don't invite that call.

Step 6: Document the Failure and Build Your Spares Inventory

Once things are stable, take an hour to document what happened. Root cause. Parts replaced. Total cost. Time lost. This isn't busywork—it's how you justify the emergency spend and prevent a repeat.

I also recommend keeping critical spares on the shelf:

• Capacitors and contactors for your AC compressors
• One spare temperature sensor and one spare controller
• A spare power supply
• Radiator keys, if you have hydronic heat

Spares are a hard sell because they cost money with no visible return. But in the next emergency, a $150 power supply on the shelf can save you a $500 overnight order and 12 hours of waiting. Trust me on this one.

Common Mistakes That Make Emergencies Worse

A few things I've seen go wrong more times than I can count:

• Leaving a buddy heater unattended. Fire risk, CO risk, insurance risk. Don't do it.
• Bleeding radiators while the system is hot. Wait. The radiator will still be there in the morning, and so will your skin.
• Buying suspiciously cheap "genuine" parts. If someone's offering an Omron power supply at 60% off retail, it's almost certainly counterfeit. You wouldn't trust a discounted Omron compressor nebulizer NE-C801 for a patient's respiratory therapy—the same logic applies to a building's critical electrical hardware.
• Cranking the thermostat to maximum after a repair. That doesn't speed anything up. It causes overshoot, short-cycling, and another trip. Set it to setpoint and let the controls work.
• Trying to save money with standard shipping on an emergency need. Five days of downtime plus a likely second rush order costs way more than paying for next-day the first time. The lowest quote isn't the lowest total cost.

Operating principle: systems will fail on holidays, at 5 PM, and during your busiest week. If you've already walked through this checklist before the failure, you're ahead of most teams. The work is straightforward—just do it in the right order.