Leak Detection Methods

For homeowners

Refrigerant leaks are the most common cause of HVAC system failure over time. A perfectly assembled system stays charged indefinitely, but real systems develop leaks at fittings, brazed joints, evaporator coils, and Schrader cores. When a system has lost charge and won’t hold pressure, finding the leak is the job before any repair.

There is no single best method. Each method has different sensitivity, different cost, and different best applications. A skilled tech uses several methods together depending on the situation.

The five common methods:

Soap bubble test. Apply leak-detection soap solution to suspect joints. Watch for bubbles forming where gas escapes. Cheap, simple, visual. Best for medium leaks at accessible joints. Sensitivity: roughly 1 ounce per year — won’t find the smallest leaks.

Electronic sniffer. Handheld instrument with a probe that’s waved near joints. Detects halogenated refrigerants in the air via a heated diode, corona discharge, or infrared sensor. Modern sniffers detect very small leaks (0.05–0.1 oz/year). Won’t detect pure nitrogen, so doesn’t work during a nitrogen-only pressure test.

Ultrasonic detector. Microphone-based instrument that picks up the high-frequency sound of escaping gas. Works with any pressurized gas including pure nitrogen. Good for finding leaks behind insulation or in hard-to-reach places.

UV dye with fluorescent lamp. A fluorescent dye is injected into the system. The dye circulates with refrigerant and oil. At a leak point, the dye comes out and stays visible under a UV lamp for weeks. Best for intermittent leaks that don’t show during a single service visit.

Pressure decay test. Pressurize the system with nitrogen and watch the gauge over time. Pressure drop indicates a leak somewhere. Doesn’t locate the leak — just confirms it exists. See nitrogen pressure test.

Common leak locations:

• Schrader cores — always check first
• Flare connections
• Brazed joints, especially at coil headers
• Evaporator coils — formicary corrosion in newer aluminum coils is a common issue
• Old line sets at fittings

Florida note: Coastal salt air accelerates copper corrosion. Outdoor coils within a few miles of the ocean develop pinhole leaks faster than the same coil installed inland. UV-resistant coil coatings help but don’t eliminate the problem.

---

For technicians

Soap bubble test.

Sensitivity: roughly 1 oz/year minimum. Smaller leaks won’t push enough gas to break the surface tension and form visible bubbles.

Equipment: commercial leak detection solution (specifically formulated for low surface tension and slow drying), or dish soap and water mixed 50/50, applied with a small brush or squeeze bottle. The commercial solution is preferable — it stays in place longer and produces visible bubbles at smaller leaks.

Strengths: cheapest method. Visual confirmation. Pinpoints the exact location.

Weaknesses: won’t find small leaks. Won’t find leaks behind insulation or inside enclosed components. The soap residue must be cleaned off after testing — left on, it can promote corrosion over time.

Best applications: post-brazing checks, post-flare-connection checks, verifying that a found leak is the only leak.

Electronic refrigerant sniffer.

Sensor types:

• Heated diode (most common today) — sensitive, specific to halogenated refrigerants, responds quickly. Sensors have a limited life (1,000–2,000 hours of operation) and must be replaced periodically.
• Infrared (IR) (modern premium) — very sensitive, very specific, long sensor life. Less affected by ambient humidity and contaminants than heated diode.
• Corona discharge (oldest, cheapest) — produces false positives from many other sources (soap, cleaning solvents, gasoline vapor). Largely phased out.

Sensitivity: 0.05–0.5 oz/year depending on instrument quality.

Procedure:

1. Calibrate the instrument by zeroing it in clean air
2. Move the probe slowly (about 1 inch per second) close to (within 1/4 inch of) suspect joints
3. When alarm triggers, mark the location and verify by moving probe away and back

Tips:

• Refrigerant is heavier than air; it sinks. Probe below the joint, not above.
• Outdoor wind disperses leaked refrigerant; shield the area with cardboard on windy days
• Modern sniffers have “auto-zero” that can mask a slow uniform leak; manually re-zero in clean air periodically

Weaknesses: requires refrigerant in the system. Sensors degrade and need replacement. False positives from cleaning agents, gasoline fumes.

Best applications: general leak finding when system is charged or partially charged. The standard tool for service tech work.

Ultrasonic leak detector.

Method: a directional microphone tuned to ultrasonic frequencies (typically 30–50 kHz) that picks up the sound of gas escaping through a small orifice.

Strengths: works with any pressurized gas — refrigerant or nitrogen. Works in noisy environments because the listening frequency is above normal noise. Can find leaks behind insulation, inside cabinets, in tight spots.

Weaknesses: requires training to interpret the sounds. False indications from other ultrasonic sources (arcing electrical, friction). Doesn’t work as well in windy conditions.

Best applications: nitrogen pressure tests (since sniffers don’t work), leaks in hidden or hard-to-access locations, large multi-zone systems where scanning many joints is faster ultrasonically.

UV dye method.

Method: fluorescent dye is added to the refrigerant system and circulates throughout with the refrigerant and oil. At a leak point, dye and oil seep out and remain on the surface. A UV lamp (typically 365–405 nm) reveals the dye as bright greenish-yellow fluorescent stains.

Sensitivity: very small leaks become detectable given time. Many sources rate UV dye sensitivity as comparable to good electronic sniffers, but the dye accumulates over hours to days, so very slow leaks reveal themselves where other methods would have missed them in a single visit.

Procedure for finding an existing leak:

1. Inject the dye into the system through a service port with a dye injector tool
2. Run the system normally for several hours to a few days (depending on suspected leak rate)
3. Return with a UV lamp and UV-blocking glasses
4. Scan the system with the lamp in low ambient light
5. Stained areas indicate leak points

Strengths: persistent indicator. Excellent for intermittent leaks. Even very small leaks eventually leave visible dye. Easy to interpret once stained areas are found.

Weaknesses: requires waiting period for dye to circulate and develop at leak point. Some manufacturers void warranty if non-approved dyes are used. Dye can interfere with refrigerant analyzers used by reclaimers.

Best applications: intermittent leaks (system loses charge but doesn’t leak when tested), small leaks below sniffer threshold, factory-fill on new installations as preventive measure.

Common leak locations in residential systems.

Schrader cores. The most common source of slow leaks. The rubber seal in the core ages and develops weeping leaks. Cheap to replace ($1 each), takes 30 seconds with a core removal tool. Always check first when investigating a charge loss.

Flare connections. The flared copper end can be over-tightened (cracks the flare), under-tightened (incomplete seal), or made on a damaged tubing end. Especially common in line sets at mini-split installations.

Brazed joints. Pinholes from incomplete penetration during brazing, or cracks from thermal stress. Often appear at coil headers and at line set connections.

Evaporator coils. Modern high-efficiency evaporator coils often use thin-wall aluminum tubing. Some designs are susceptible to formicary corrosion — microscopic pinhole leaks caused by formic acid attacking the aluminum from inside (from formic acid present in some indoor air contaminants like fresh paint, new carpet glues, and certain cleaning products). Has been a significant warranty issue in residential systems since around 2010.

Outdoor coils. Mostly from physical damage (lawnmower debris, weed trimmer line, fin damage from yard maintenance). Less commonly from coastal salt corrosion.

Service caps. Loose caps allow Schrader leaks to vent to atmosphere. Always replace caps after service.

Diagnostic workflow when a system has lost charge.

1. Check Schrader caps and cores first. Tighten caps. Replace cores. Many “leaks” are simply Schrader weeps.

2. Visual inspection of accessible joints and components. Look for oily residue — refrigerant oil leaks out with refrigerant and marks the leak location.

3. If refrigerant remains, use electronic sniffer on all accessible joints. Spend more time at known weak points.

4. If no refrigerant remains or system can’t be charged for testing, perform nitrogen pressure test plus ultrasonic detection. Or pressurize with nitrogen + trace charge of refrigerant for sniffer testing.

5. If leak is intermittent, inject UV dye and return after a few days.

6. For coil-internal leaks, the coil must often be removed and pressure-tested separately, or replaced empirically when other methods can’t locate the leak.

7. Document findings and quote repair or replacement. Some leaks (Schrader cores, accessible brazed joints) are cheap to repair. Coil leaks often require coil replacement, which on older systems may not be economical compared to full system replacement.

Florida-specific leak issues.

• Coastal salt corrosion: outdoor units within 5 miles of the ocean see accelerated corrosion. Coil coatings (electrocoat, e-coat) extend life but don’t eliminate the problem. Salt-corroded coils typically last 5–10 years versus 15–20+ inland.
• Attic coil installations: indoor coils mounted in unconditioned attics see large temperature swings. Thermal cycling stresses brazed joints. Coil failures from thermal fatigue are more common in attic installs.
• High humidity: indoor coils run wet for months at a time. Combined with formicary corrosion susceptibility, this is the worst-case environment for aluminum coil failure.