As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
Grams per year is a mass flow: the quantity of gas escaping from a leak over a period of time, not the concentration the instrument sees in the air. A detector described as sensitive to 4 grams per year is claiming it can find a leak losing that much mass annually, under the specific conditions its manufacturer used. Two such figures only compare when the gas, the air movement around the source and the way the probe was moved are the same in both cases, which is why the number should always be read together with the conditions attached to it.
A gram per year is a mass flow, not a concentration
Two different quantities get muddled under the word sensitivity. One is concentration: how much gas is present in a sample of air, expressed in parts per million or as a percentage by volume. That is what a fixed monitor in a plant room measures and what an alarm threshold is set against. The other is leak rate: how much mass escapes from an opening per unit time, expressed in grams per year or ounces per year. That is what a handheld sniffer is specified against.
The two are connected but not interchangeable. A leak of a given mass flow produces a plume whose concentration depends on how fast the gas disperses once it leaves the opening, which depends on air movement, on the geometry around the joint and on whether the gas is heavier or lighter than air. Put the same calibrated leak in a sealed cupboard and in a windy plant room and the concentration a probe can find at the same distance differs by orders of magnitude, while the mass flow has not changed at all.
That is why the mass flow figure is the honest one to specify an instrument against, and also why it can only be established under controlled conditions. The figure describes a capability demonstrated in a defined situation, and the situation is part of the specification. Where a listing publishes the number without the situation, the number is an aspiration. The sensitivity explainer covers how that plays out across detector types.
The three conditions that make a sensitivity figure comparable
First, the gas. A heated diode responds to halogen atoms and will not see a hydrocarbon refrigerant at all. An infrared bench is tuned to an absorption band and responds differently to every molecule that absorbs near it. A figure quoted for one refrigerant tells you very little about another with a different molecular structure, and nothing whatsoever about a refrigerant the sensing principle cannot detect.
Second, the air. The published figures come from still air, usually inside an enclosure, because any draught thins the plume to nothing within a short distance of the opening. Real plant rooms have ventilation, open doors and running fans. A figure proven in still air is a ceiling on what the instrument will achieve in a space with air movement, not a prediction of it.
Third, the handling. Probe distance from the opening and the speed at which the probe is moved both dominate the result, because the sensor needs time in the plume to respond. Standards for refrigerant leak detectors fix these: a defined distance, a defined sweep rate, a defined response the instrument must produce. That is what separates a specification from a claim, and it is the detail most worth looking for when choosing a refrigerant leak detector.
Converting between ounces per year and grams per year
One avoirdupois ounce is 28.35 grams, so the conversion is a single multiplication and the figures quoted in listings line up predictably once converted. The rounding in the right hand column below is the manufacturers own, which is worth knowing when two instruments appear to differ and in fact do not.
| Ounces per year | Grams per year, exact | How listings usually round it |
|---|---|---|
| 0.03 | 0.85 | Under one gram per year |
| 0.05 | 1.42 | 1.5 grams per year |
| 0.1 | 2.83 | 3 grams per year |
| 0.14 | 3.97 | 4 grams per year |
| 0.15 | 4.25 | 4 grams per year |
| 0.5 | 14.17 | 14 grams per year |
The Inficon Compass states a sensitivity of 0.15 ounces per year, which it gives as 4 grams per year, together with a heated diode sensing element, detection of CFCs, HCFCs, HFCs and HFOs, and a claim of no false alarms from soap solutions. That is a complete enough statement to argue with: a figure, both units, a sensing principle and a named refrigerant family.
What four leak detector listings publish about sensitivity
Put four products side by side and the variation in disclosure is larger than the variation in capability. The Robinair LD7 describes a colour graphic display with bar graph and sweep modes, colour change as the sensor approaches the leak, an audible alarm, ultraviolet LEDs for dye work and an inspection light. All of that is about finding the leak once you are near it, and a sensitivity figure is not stated in the listing bullets. The Testo 0632 0336 gas leak detector carries no specification bullets at all, so sensing principle, target gases and sensitivity are all unstated and have to come from the manual.
| Instrument | Sensitivity stated | Gas or family named | Quantity it is specified against |
|---|---|---|---|
| Inficon Compass refrigerant leak detector | 0.15 ounces per year, given as 4 grams per year | CFCs, HCFCs, HFCs and HFOs | Mass flow from a leak |
| Robinair LD7 refrigerant leak detector | Not stated in the listing bullets | Not stated in the listing bullets | Mass flow, by instrument class |
| Testo 0632 0336 gas leak detector | Not stated in the listing bullets | Not stated in the listing bullets | Not stated in the listing bullets |
| Plug in natural gas and carbon monoxide alarm | Not stated in the listing bullets | Natural gas, qualitatively | Concentration in the room, not mass flow |
The last row is the one that trips people up. The plug in natural gas and carbon monoxide alarm states that it senses dangerous levels of natural gas, emits an audible and visual alert, installs in any standard outlet and carries a ten year sealed battery, with a sensor described as responsive to trace amounts. It is a room monitor and belongs to a different specification language entirely, one built around alarm concentrations rather than leak rates. Comparing its sensitivity with a sniffer is comparing a smoke alarm with a torch.
Using a sensitivity figure without being misled by it
Convert everything to one unit before comparing, then ask which gas each figure belongs to and discard any that do not name one. Ask whether the figure is a detection threshold or an alarm threshold, because an instrument that twitches at a level is not the same as one that announces it. Then work out whether the threshold is relevant to the systems you service: calculate the loss as a percentage of the charge rather than treating the mass alone as large or small.
| Charge in the system | Annual loss at 4 grams per year | Time to lose a tenth of the charge |
|---|---|---|
| 60 grams, small sealed appliance | About 6.7 percent | Around eighteen months |
| 500 grams, vehicle air conditioning | About 0.8 percent | Around twelve years |
| 1500 grams, residential split system | About 0.27 percent | Around thirty seven years |
Those three rows explain why the same instrument is overspecified for one job and barely adequate for another. Finish by confirming the instrument can actually be verified in the field, because a sensitivity figure that cannot be rechecked drifts quietly: a calibrated reference leak is the only honest way to test a gas leak detector on a bench. For the wider picture of what this class of instrument does and does not do, start with the refrigerant leak detector explainer.
Frequently Asked Questions
Does a smaller grams per year figure always mean a better detector?
Not on its own. A lower threshold is genuinely useful when you are hunting a slow loss that a soap test will never show, but a very sensitive instrument in a space with any residual vapour will respond everywhere and pinpoint nothing. That is why instruments offer selectable sensitivity levels rather than running at maximum. The useful figure is the lowest level at which the instrument can still discriminate between a joint and the air around it.
How is a leak rate in grams per year actually established?
With a calibrated leak source: a reference device that emits a known mass of a named gas per unit time, verified by weighing or by a certified permeation rate. The source is placed in still air in an enclosure, the probe is brought to it in a defined way, and the response is recorded. Everything that varies between laboratories, principally air movement and probe handling, is fixed by the procedure, which is why two figures from different procedures are not comparable.
Does a sniffer tell me how big a leak is?
No. A sniffer reports that vapour is present near the probe tip, usually on a relative scale, and the reading depends as much on how close the tip is and which way the air is moving as on the size of the leak. Quantifying a loss needs a different method: weighing the charge, recovering and measuring it, or enclosing the component and letting vapour accumulate for a measured time. Treat a sniffer as a locator.
Why do some listings give ounces and some give grams?
Because the automotive service trade in some markets standardised on ounces per year while refrigeration and the international standards work in grams per year, and manufacturers publish whichever suits the audience. The conversion is fixed, since one ounce is just over 28 grams, so the two scales are interchangeable with arithmetic. What is not interchangeable is the test condition behind each figure.
Is a fixed alarm specified the same way?
No, and confusing the two leads to meaningless comparisons. A fixed or plug in alarm is specified by the concentration at which it alarms, expressed in parts per million or as a percentage of the lower explosive limit, because its job is to warn when a room has become hazardous. A sniffer is specified by the smallest mass flow it can find at a joint. One measures the state of a room, the other searches a surface.
How long does a small leak take to matter?
It depends entirely on the charge, which is why the same annual mass loss is trivial on one system and serious on another. A loss of a few grams a year is a rounding error against a large commercial charge and a significant proportion of a small sealed appliance. Work the percentage before deciding whether a figure is alarming, because the mass flow on its own carries no sense of scale.