This is a paper I found and had a copy of,can't tell you the author but this was written around 30+ years ago.
System Residues hide leaks. Sludge and residues coat the interior of system components and temporarily seal corrosion pits, fissures, seams, seals, o’rings, and other small leak points. Some of those residues include refrigeration oils, acids, desiccant, pulverized metal, Teflon piston ring material, brazing fluxes, dye particles, etc.
Overcoming Residue Surface Tension: Leaks are harder to find because leak testing is performed with the system turned off. Lost are the benefits of an operating system: a) constant washing of interior surfaces of components; b) higher operating pressures that encourage leaks. With a system at rest, the undisturbed residues mentioned above are able to coat the insides of the evaporator, condenser, compressor, and other components. When Residue Surface Tension is greater than the interior pressures’ ability to displace it, there is no leak occurrence, thus no leak detection.
Residue Displacement: If surface tension is the culprit, then how do we overcome it? Answer: By adding 4-ounces of chlorine-based refrigerant R-22 to the system. R-22 disturbs the surface tension. We follow that by pressurizing the system with 175-200 psi of nitrogen. Both are cheap.
At the higher system pressure, the R-22 overcomes the residue surface tension and forces the leak path to reopen. Now, enough R-22 gas is available so that leaks are detected easily. The R-22 is the residue-displacing agent.
R-22 summary. R-22: (a) creates a leak path because of its oil solubility and residue displacement potential; (b) is more readily detectible by electronic leak detectors than HFCs; (c) is non visual-dependent, unlike dyes, thus can be used effectively to locate leaks in enclosed areas; (d) will not cross-contaminate nor cause any harmful consequences to CFC, HFC, nor blend refrigerant systems (i.e., after leak testing an HFC (R-134a) system using R-22, there is typically zero percent cross-contamination if evacuated afterwards).
Nitrogen summary. Nitrogen has the following qualities: (a) inert, very dry and non-flammable; (b) does not go into solution with refrigeration oil to create non condensable pressure problems (i.e., after leak testing an HFC (R-134a) system using R-22 and nitrogen, there is typically zero percent non condensables (nitrogen or air), if evacuated afterwards).
Diluted Liquid Soap. Speed is important to technicians when performing a leak test. Large and medium size leaks can be quickly located in exposed areas such as under hood components by applying diluted liquid dishwashing soap directly to suspected leak points.
Multiple Leaks. If a large leak is discovered repair it and perform a follow-up leak test. Leak tests should be performed following each repair until all leak points are found and fixed.
Pinpoint versus Area Testing Pinpoint Accuracy. Technicians need to know exactly where leaks exist. Consequently, leak testing needs to be pinpoint accurate, regardless whether by visual or non-visual means. With dyes, the detection is strictly visual and general area, not pinpoint. Dyes do not provide the rapid and finite definition of the Combination Method [non-visual: electronic leak detector detecting the nitrogen-pressured R-22, and, visual: diluted soap solution bubbling].
Electronic Leak Detectors. Heated diode leak detectors are currently the best non-visual, dependent leak detection instrument for sensing a gaseous leak. They have the necessary characteristics of sensitivity, repeatability and recoverability (after a leak), which are so important.
Summary
There is currently no equally effective (in all situations) substitute for the combination leak test method described above. It has neither the problems nor shortcomings of dye. I have heard the argument that nitrogen is dangerous. Well, the stationary industry has been using it for at least 60 years, and my company has been using it since 1973. Some of the major stationary manufacturers think that is important enough to dedicate a section of their training manuals exclusively to its use for leak testing their a/c systems. It is not unsafe, unless one uses it improperly. Mobile a/c technicians are simply untrained in its use. What a shame, it is so simple and effective and easier, safer, and cheaper, etc.
Once upon a time, folks were afraid to fly in an airplane that didn’t have a propeller—it was thought to be too dangerous. I encourage any a/c technician to get a good quality nitrogen regulator and bottle of nitrogen from a welding supply store. Set the regulator to 200 psi. It is safe to use with all conventional a/c systems, and will not damage any system components.
Make sure that the nitrogen bottle is secured to a wall or in a dolly, and that everyone in the service department is taught not to fool with the regulator setting. Or, install an Allen head set screw and lock nut with the regulator so that it can’t be adjusted once it is pre-set. We have safely tested a/c system components at 200-psi nitrogen pressures on over 40,000 tests. Remember, 200 psi is only about 50 psi higher than the compressed air lines used by you and every other technician nation wide. I encourage all technicians who aspire to become more professional in a/c service to invest in a good quality nitrogen regulator and bottle. Then take the time to practice the method. You will be doing a service to yourself and your customers. $$$
I think he was a pilot
I forgot "you must" 
Reply With Quote
is the remedy 
Originally Posted by runmc
) it works well and maybe it has the same capacity than R22 for cleaning. That would be interesting to know a bit more about it
Bookmarks