Figure 1. In leaking systems, oil spots can appear wet and have a fine coating of dust. (Feature photos courtesy of Refrigeration Technologies.)

Every environmentally conscious service technician should spend time learning how to check for refrigerant leaks in refrigeration and/or air conditioning systems. Ozone depletion, global warming, and the increasing price of refrigerants are forcing technicians to become better and more thorough leak detectors. This column will cover some basic methods of leak detection in refrigeration and air conditioning systems. My Nov. 7 column will look at some of the more advanced methods.


All sealed systems leak. The leak could be 1 pound per second or as low as 1 ounce every 10 years. Every pressurized system leaks because “flaws” exist at every joint fitting, seam, or weld. These flaws may be too small to detect with even the best of leak detection equipment. But given time, vibration, temperature, and environmental stress, these flaws become larger detectable leaks.

It is technically incorrect to state that a unit has no leaks. All equipment has leaks to some degree. A sealed system which has operated for 20 years without ever needing a charge is called a “tight system.” The equipment still has leaks, but not enough leakage to read on a gauge or affect cooling performance. No pressurized machine is perfect.

A leak is not some arbitrary reading on a meter. Gas escapes at different times and at different rates. In fact, some leaks cannot be detected at the time of the leak test. Leaks may plug, and then reopen under peculiar conditions. A leak is a physical path or hole, usually of irregular dimensions. The leak may be the tail end of a fracture, a speck of dirt on a gasket, or a microgroove between fittings.


Refrigerant vapor can flow under layers of paint, flux, rust, slag, and pipe insulation. Often the refrigerant gas may show up quite a long distance from the leak site. This is why it is important to clean the leak site by removing loose paint, slag, flux, or rust. Remove any pipe insulation. Oil and grease must also be removed from the site because they will contaminate the delicate detection tips of electronic detectors.

There are six classifications of leaks.

1. Standing leaks:Standing leaks are leaks that can be detected while the unit is at rest or off. This includes freezer evaporator coils warmed up by defrost. Standing leaks, fortunately, are the most common of all leaks.

2. Pressure-dependent leaks:Pressure-dependent leaks are leaks that can only be detected as the system pressure increases. Nitrogen is used to pressurize the low sides of systems to around 150 psig, and high sides to 450 psig. Never use air or pure oxygen. Often, a refrigerant trace gas is introduced into a recovered and evacuated system along with the nitrogen. The trace gas enables electronic leak detectors to be used to detect the vicinity of the leak. Refrigerant trace gas will be covered in more detail later in the article. Pressure-dependent leak testing should be performed if no leaks are discovered by the standing leak test. Bubbles or a microfoam solution can also be used to locate pressure-dependant leaks.

Warning: Mixtures of nitrogen and a trace gas of refrigerant, usually of the system’s refrigerant, can be used as leak test gases because, in these cases, the trace gas is not used as a refrigerant for cooling. However, a technician cannot avoid recovering refrigerant by adding nitrogen to a charged system. Before nitrogen is added, the system must be recovered and then evacuated to appropriate levels. Otherwise, the HCFC, or HFC refrigerant trace gas vented along with the nitrogen will be considered a refrigerant. This will constitute a violation of the prohibition on venting. Also, the use of CFC as a trace gas is not permitted.

3. Temperature-dependent leaks:Temperature-dependent leaks are associated with the heat of expansion. They usually occur from high-temperature ambient air, condenser blockages, or during a defrost period.

4. Vibration-dependent leaks:Vibration-dependent leaks only occur during unit operation. The mechanical stain of motion, rotation, refrigerant flow, or valve actuation are all associated with vibration-dependent leaks.

5. Combination dependent leaks:Combination dependent leaks are flaws that require two or more conditions in order to induce leakage. For example, temperature, vibration, and pressure cause the discharge manifold on a semi-hermetic compressor to expand and seep gas.

6. Cumulative microleaks:Cumulative microleaks are all the individual leaks that are too small to detect with standard tools. The total refrigerant loss over many years of operation slightly reduces the initial refrigerant charge. A system having many fittings, welds, seams, or gasket flanges will probably have a greater amount of cumulative microleaks.

Figure 2. Oil is always present at Schrader valves and access ports due to the discharging of refrigerant hoses on the manifold gauge set.


Successful leak detection is solely dependent on the careful observation made by the testing technician. Fortunately, all refrigeration systems internally circulate compressor oil with the refrigerant. Oil will blow off with the leaking refrigerant gas and “oil mark” the general area of leakage. Oil spots appear wet and have a fine coating of dust (Figure 1). The technician must determine that the area wetness is oil and not condensate. This can be accomplished by rubbing the area with your fingers and feel for oil slickness. However, what is the reliability of oil spotting? Oil spotting is the technician’s first quick-check, but is not always reliable for the following reasons:

• Oil is always present at Schrader valves and access ports due to the discharging of refrigerant hoses on the manifold and gauge set. (Figure 2). Often these parts are falsely blamed as the main point of leakage.

• Oil blotches can originate from motors, pumps, and other sources.

• Oil residue may be the result of a previous leak.

• Oil is not always present at every leak site. It may take months, even years of unit operation to cause enough oil blow-off to accumulate on the outer side.

• Oil may not be present with micro-leaks.

• Oil may not reach certain leak positions.

• Oil may not be present on new start-ups.


Many leaks that go undetected are in the evaporator coil. This is because most evaporator sections are contained in cabinets, buttoned-up, or framed into areas that do not allow easy access. In order to avoid time-consuming labor to strip off covers, ducting, blower cages, or the unloading of product, an easy electronic screening method is outlined below.

1.Turn off all system power including evaporator fan motors.

2.Equalize high- and low-side pressures in the refrigeration or air conditioning system and defrost any frozen evaporator coils. (If the system does not have any pressure, evacuate to the required levels and then add a refrigerant trace gas. Nitrogen is then added to generate a practical test pressure). Most low sides of systems have a working pressure of 150 psig, but always read the nameplate on the evaporator section for test pressure specifications.

3.Calibrate an electronic leak detector to its highest sensitivity.

4.Locate the evaporator drain outlet or downstream drain trap.

5.Position the electronic leak detector probe at the drain opening. Be careful that the leak detector probe does not come into contact with any water.

6.Sniff with the electronic detector for a minimum of 10 minutes or until a leak is sensed. Recalibrate the device and test again. Two consecutive “positive” tests confirm an evaporator leak. Two consecutive “negative” tests rules out an evaporator section leak.

Remember, refrigerant gas is heavier than air. Gravity will cause the gas to flow to the lowest point. If the evaporator section tests positive, the technician should expose the coil and spray coat all surfaces with a specially formulated bubble/foam promoter. Bubble/microfoam solutions have been very successful in leak detection because of their price and effectiveness. Leaks can be easily pinpointed with these solutions. Often, a mild soap and water solution is used for bubble checking. Research has shown that soap and water does not have the same properties as do the micro-foam solutions that contain coagulants and wet adhesives. Household detergents often contain chlorides and will pit and corrode brass and iron.

The specially formulated and patented bubble solutions have entered the market with remarkable results. These new solutions will form a foam “cocoon” when in contact with a leak. All that is required is for the solution to be applied over the suspected leak area. When a leak is found, bubbles or foam will tell the technician of its location. The technician must be patient and let the bubbles stand for at least 10-15 minutes if small leaks are suspected.

Another advantage of bubble testing is that bubbles can be used with nitrogen or refrigerants pressurizing the system. Small, significant leaks of less than a couple ounces per year can often be found with special formulated microfoam solutions. If the evaporator section tests negative for leaks, continue on to leak testing the condensing unit.

Figure 3. In testing for leaks, cover the condensing unit with a cloth tarp or bed sheet to serve as a barrier against any outside air movement and to trap refrigerant gas.


To test for condensing section leaks:

1.Calibrate an electronic leak detector to its highest sensitivity and place the probe at the base of the unit, usually under the compressor. The unit should be fully pressurized.

2.Cover the condensing unit with a cloth tarp or bed sheet to serve as a barrier against any outside air movement and to trap refrigerant gas (Figure 3). Do not use a plastic material because some plastics may set off some electronic leak detectors and give a false reading.

3.Monitor for leakage for 10 minutes or until a leak is sensed. Recalibrate and test again. Two consecutive positive tests confirm condensing section leakage. Two consecutive negative tests rule out a detectable leak.

4.Use the electronic leak detector to check for leaks on the bellows of the pressure controls. Remove the control box cover and place the probe within the housing. Cover the control tightly with a cloth barrier and monitor for 10 minutes as above.

5.If the results are positive, uncover the equipment and begin spray coating with a microfoam solution. If the results are negative, continue to the suction/liquid line leak test that follows.


The longer the tubing runs are between the evaporator and condensing unit, the greater is the odds for defects. Count on all possibilities whether it be a typical sight glass-drier connection leak to a poor solder joint hidden under pipe insulation.

The suction line can be screened by calibrating an electronic leak detector to its highest sensitivity. Tuck the probe underneath the pipe insulation. Monitor for 10 minute intervals while the system is at rest and fully pressurized to equalization. It may be necessary to insert the probe at several downstream points.

If a leak is sensed, strip off insulation and apply a bubble/foam promoter to all surfaces. If no leak was positively screened, test the liquid line.

Publication date:10/04/2010