As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
They are not competing methods. Refrigerant sniffing, using an instrument such as the Inficon Compass heated diode detector, finds leaks in a system that still holds refrigerant by sensing halogen gas in the air around it. Tracer gas detection is used on an evacuated or empty circuit: you charge a known gas, usually a nitrogen and hydrogen mix, and search for where it escapes. If the system is charged and running, you want a refrigerant sniffer. If the circuit is empty, under repair, or you need to prove a joint is sound before charging, you want tracer gas. The two are sequential steps in a repair, not alternatives.
What each method physically measures, and why that decides the job
A refrigerant sniffer is a gas detector tuned to halogenated compounds. It draws air across a sensor and responds when the concentration of refrigerant vapour rises above the background. The sensor does not know where the gas came from. It only knows that the air passing it contains more of the target molecule than the air a moment ago. That is why a sniffer can find a leak in a charged system without any preparation: the refrigerant is already there, escaping.
Tracer gas detection works on the opposite principle. You remove the refrigerant, evacuate the circuit, and charge it with a gas you can detect cheaply and safely, most often a mixture of nitrogen and a small percentage of hydrogen. You then probe the joints, brazed connections and coil faces with a detector tuned to that gas. The detector responds to hydrogen, not to refrigerant. The circuit is under pressure, so gas escapes at the leak and the probe finds it.
The consequence is simple. A refrigerant sniffer needs refrigerant in the system. A tracer gas method needs the system empty of refrigerant and filled with the tracer. You cannot swap one for the other without changing the state of the system first. That single fact decides most of the choice before sensitivity, response time or probe design are considered.
The measurement units differ too. A refrigerant sniffer responds to concentration in air, often expressed in parts per million, and some instruments also offer a percentage of the lower explosive limit for flammable refrigerants. That second scale matters for A2L refrigerants, which are mildly flammable, and it is a different question from leak location. The difference between ppm and LEL is worth understanding before you read a dual-scale instrument as two sensitivities of the same thing.
Reading a refrigerant sniffer specification without being misled
Refrigerant sniffer listings quote sensitivity in different ways and the numbers are not directly comparable. The Inficon Compass listing states a sensitivity of 0.15 ounces per year, which is about 4 grams per year. That is a leak rate, not a concentration. It means the instrument is claimed to respond to a leak losing that mass of refrigerant over a year, under the manufacturer’s test conditions. It is a useful figure because it maps onto the practical question: will this find a leak small enough to matter before the system loses its charge over a season.
Other listings do not publish a leak rate at all. The Bacharach H-10 Pro listing gives the product number and category but no sensitivity figure, so the honest position is that the sensitivity is not stated in the listing bullets. The Robinair LD7 listing describes a colour display with bar graph and sweep modes but does not publish a leak rate or a detection threshold. The Aprvtio WJL-6000PRO listing describes dual ppm and percentage LEL modes and an all-metal probe but does not publish a sensitivity figure either. Where a number is missing, do not assume it matches the one that is published.
Display resolution is not sensitivity. A detector that shows a reading to one part per million is not necessarily able to find a one part per million leak. The sensor’s limit of detection and the display’s resolution are separate specifications, and consumer listings often quote the more impressive-looking one. If a listing quotes an extremely fine figure for a handheld refrigerant sniffer, check whether it is describing the display rather than the sensor.
The sensor type matters for what the instrument can see. The Inficon listing names a heated diode sensor, which responds to halogenated refrigerants including CFCs, HCFCs, HFCs and HFOs. Heated diode sensors are sensitive but can be poisoned by exposure to large concentrations of refrigerant or to contaminants, and they have a finite life. That is a maintenance cost, not a defect. Infrared and other sensor types behave differently. The listing tells you which sensor is inside; it rarely tells you how long it lasts.
Where tracer gas wins, and where it is the only honest answer
Tracer gas is the right method when the system is empty. If a circuit has lost its charge, been opened for repair, or is being assembled, there is no refrigerant to sniff. You can pressurise with dry nitrogen and listen, but nitrogen tells you nothing about a small leak unless you watch a gauge for a long time. Adding a trace of hydrogen to the nitrogen gives a detector something to find. The detector responds to hydrogen at the leak point, and because hydrogen molecules are small they pass through openings that larger molecules might not, which is why the method is used for proving joints and for finding leaks that a pressure test alone would take hours to reveal.
Tracer gas is also the method for proving a repair. After brazing a joint, you want to know the joint is sound before you evacuate, charge and leave. A tracer gas check on the repaired section is a direct test of that joint. A refrigerant sniffer cannot do this on an empty circuit, and a soap bubble test only shows you the largest leaks. The bubble test still has a place, as the tracer gas method explained sets out, but it is a coarse tool compared with a detector tuned to a tracer.
The limits of tracer gas are practical. You need a source of the tracer gas, a regulator and a way to introduce it safely. You need to evacuate the system first, which means a vacuum pump and time. You need to ventilate the area because you are releasing a gas, even a safe one, into a workspace. And you need a detector calibrated for the tracer you are using, not for refrigerant. None of this is difficult, but it is a procedure, and procedures take longer than picking up a sniffer and walking around a running system.
Tracer gas also does not tell you how much refrigerant the system has lost or whether the leak is the only one. It locates a leak point. The diagnosis of why the system is low on charge is still yours to make.
Where refrigerant sniffing wins, and what it cannot tell you
Refrigerant sniffing is the right method when the system is charged and you need to find the leak without recovering the charge. That is the common case in service work. A system is cooling poorly, the pressures suggest a low charge, and the customer wants it fixed. Recovering the refrigerant to perform a tracer gas test would add a recovery, an evacuation and a recharge to every leak call. A sniffer lets you search first.
The instrument design affects how you search. The Robinair LD7 listing describes a bar graph mode that changes colour as the sensor approaches the leak and a sweep mode that shows a line graph so you can trace back and forth to pinpoint the source. That is a real usability difference: a leak detector that gives you a trend, not just a beep, lets you triangulate. The Aprvtio WJL-6000PRO listing describes an audible tone that changes to a siren and a mute function, which matters in a quiet indoor environment where you need to hear the tone change rather than a fixed alarm. The Bacharach H-10 Pro is a long-established refrigerant detector design, but the listing supplied here does not publish its sensitivity or sensor type, so treat those as not stated in the listing bullets.
What a refrigerant sniffer cannot do is prove a system is leak-free. It finds leaks above its detection threshold in the places you probe. A leak hidden behind insulation, inside a wall, or on a section you did not reach will not be found. A negative result means you did not find a leak, not that none exists. That distinction matters when you are deciding whether to recharge and leave or to keep looking.
A refrigerant sniffer also does not measure the size of the leak in any unit you can act on directly. The bar graph and the tone tell you the concentration is rising as you approach. They do not tell you the leak rate in grams per year. The Inficon listing’s 0.15 ounces per year figure is a sensitivity claim, not a measurement the instrument reports in use.
The dual-mode instrument and the A2L question
Some instruments now offer both a refrigerant detection mode and a percentage LEL mode. The Aprvtio WJL-6000PRO listing describes exactly this: ppm for micro leak detection and percentage LEL for monitoring A2L refrigerant safety. These are two different jobs in one housing.
The ppm mode is a leak-finding mode. It responds to small concentrations of refrigerant in air and helps you locate a leak point. The percentage LEL mode is a safety mode. It responds to the concentration of a flammable gas as a fraction of the lower explosive limit, which is the concentration below which the gas and air mixture is too lean to ignite. For A2L refrigerants, which are mildly flammable, that scale tells you whether the atmosphere around you is approaching a flammable concentration. It does not tell you where the leak is with the same precision, and it is not a substitute for a leak-finding mode.
The distinction matters because a dual-mode instrument can be read as two sensitivities of the same measurement. It is not. Understanding ppm, LEL and volume is the background you need to read both scales correctly. A reading of a few percent LEL is a safety warning, not a small leak. A reading of a few ppm in leak mode is a hint that you are close, not a measure of the leak rate.
If you work with A2L refrigerants, a percentage LEL mode is a genuine safety feature and worth having. If you work only with A1 refrigerants, which are not flammable, the LEL mode adds little to leak finding and you should not pay for it as though it were a second sensitivity.
When neither method is the right answer
There are cases where the honest answer is that neither tracer gas nor refrigerant sniffing will solve the problem. If the leak is inside a sealed component such as a compressor or a moulded coil end, neither method will reach it without dismantling. If the system is losing charge so slowly that the loss is only visible over months, a sniffer may not find it in a single visit and a tracer gas test may be the only practical way to prove the circuit is sound before ruling out the system. And if the question is whether the space is safe to occupy, neither instrument is a safety monitor. A refrigerant sniffer is a leak-finding tool, not a fixed gas alarm, and it does not replace a refrigerant monitor for occupied spaces.
It is also worth being clear about what a combustible gas detector does. A combustible gas detector responds to flammable gases and is not a refrigerant sniffer. A refrigerant sniffer responds to halogenated refrigerants and is not a combustible gas detector. The two are sometimes confused because both are handheld instruments with a probe. The difference between a refrigerant leak detector and a combustible gas detector is a safety point, not a shopping point: using the wrong one can give you a false sense of security.
Finally, if you are building a kit for field work, the choice between tracer gas and sniffing is not either-or. The practical kit contains a refrigerant sniffer for charged systems and the fittings and detector for tracer gas work on empty ones. Building a gas detection toolkit is a better framing than choosing a single method for all jobs.
The decision in one paragraph
If the system holds refrigerant, sniff it. If the system is empty or you need to prove a joint before charging, use tracer gas. The instruments are not interchangeable because they measure different gases in different system states. Choose the method that matches the state of the system in front of you, and treat a negative sniffing result as a reason to keep looking rather than a proof of soundness.
Frequently Asked Questions
Can a refrigerant sniffer find a leak in an empty system?
No. A refrigerant sniffer responds to refrigerant vapour in the air. If the system has been evacuated or has lost its charge, there is nothing for it to detect. You need to charge the system with refrigerant or use tracer gas.
Is tracer gas more sensitive than a refrigerant sniffer?
Not necessarily, and the comparison is not straightforward. Tracer gas detection can find very small leaks because hydrogen molecules are small and the detector is tuned to them. Refrigerant sniffer sensitivity is quoted in different ways, often as a leak rate such as 0.15 ounces per year in the Inficon Compass listing. The two figures measure different things and cannot be ranked without knowing the test conditions.
What does the percentage LEL mode on a dual-mode detector actually measure?
It measures the concentration of a flammable gas as a fraction of the lower explosive limit, which is the concentration below which the gas and air mixture is too lean to burn. For A2L refrigerants it is a safety scale, not a leak-finding scale. A few percent LEL is a warning, not a small leak.
Does a refrigerant sniffer tell me the size of the leak?
No. It tells you that the concentration of refrigerant is rising as you approach the leak. It does not report a leak rate in grams per year or ounces per year. The sensitivity figure in a listing, where one is published, is a claim about the smallest leak the instrument can respond to, not a measurement it reports in use.
Can I use a combustible gas detector to find refrigerant leaks?
No. A combustible gas detector responds to flammable gases and is not tuned to halogenated refrigerants. A refrigerant sniffer responds to refrigerants and is not a combustible gas detector. Using one in place of the other can miss the hazard you are looking for.
Which method should I use after brazing a joint?
Tracer gas, because the circuit is empty at that point. You pressurise with a nitrogen and hydrogen mixture and probe the joint. A refrigerant sniffer cannot test an empty circuit, and a soap bubble test will only show the larger leaks.