A breaker can hold full gas density and still be unsafe to energise. Density tells you how much SF6 is in the compartment; it says nothing about what else is in there. That is why an SF6 gas purity test — together with moisture and decomposition product measurement — is the check that decides whether a compartment goes back into service.
This guide explains what the three quality parameters actually measure, which standard applies to which situation, how the measurements are made, and where sampling errors ruin otherwise good results. It covers outdoor breakers from 40.5 kV to 252 kV of the kind in the high voltage SF6 circuit breakers range. Numeric limits vary by edition and by utility specification, so always apply the current edition of the relevant standard and the breaker manufacturer’s manual — the values here are the orders of magnitude you should expect, not a substitute for either document.
The most common mistake in gas quality work is applying the wrong standard to the wrong gas. There are two questions, and they have different answers.
| Standard | Question it answers | Applies to |
|---|---|---|
| IEC 60376 | Is this gas acceptable as new gas delivered to site? | Fresh gas from a supplier, before it enters any equipment |
| IEC 60480 | Is this gas acceptable for reuse after recovery? | Gas taken out of equipment and reclaimed |
| IEC 62271-1 | Is the gas in service still fit for the equipment? | Gas inside an installed breaker, checked during maintenance |
These are not interchangeable. New gas that fails the delivery standard goes back to the supplier. Recovered gas that meets the reuse standard can go back into a breaker. In-service gas is judged against the limits for the equipment, which are usually more permissive than the new-gas limits because the compartment is sealed and the gas has already been accepted once.
| Parameter | What it measures | What a bad result means |
|---|---|---|
| Purity | SF6 content as a percentage, with air and nitrogen as the main diluents | Air ingress during filling or a leaking compartment |
| Moisture | Water content, expressed as dew point or as ppm by volume | Dielectric strength loss and acid formation with arc products |
| Decomposition products | SO2, SOF2, SO2F2, H2S and related species | Arcing has occurred; corrosive compounds are present |
| Acidity and hydrolysable fluorides | Acid-forming content, expressed as HF equivalent | Corrosion risk to internal metal surfaces and seals |
In practice, the first three carry the diagnostic weight. Acidity is usually measured on reclaimed gas at a laboratory rather than in the field.
Purity is expressed as the SF6 mass or volume fraction, with the balance being air, nitrogen and other residuals. New gas is normally specified at 99.7% or better, and reuse criteria are typically a point or two lower. Air content is the parameter that matters operationally, and it is usually specified separately because it is the one that changes when a compartment is filled without proper evacuation.
Two things purity does not tell you. It does not detect moisture, because water at the tens-of-ppm level barely moves a percentage figure. And it does not detect decomposition products, because those are also present at ppm levels. A gas sample reading 99.9% pure can still be wet and full of arc byproducts. That is why purity alone is never a release criterion.
Air ingress has a specific signature: purity and air content both move, moisture may rise slightly, and decomposition products stay clean. If you see that combination on a newly commissioned compartment, suspect the filling procedure rather than the equipment. On an in-service compartment with the same signature, suspect the leak path — air entering means gas leaving.
Moisture is the parameter most often misreported, because it is quoted in two units that are not linearly related and are not measured at the same reference condition.
Dew point is the temperature at which water vapour in the gas begins to condense at a stated pressure. It is the physically meaningful number for a breaker, because what matters is whether condensation can occur on a cold internal surface at operating pressure.
ppmv is a concentration ratio. Converting between the two requires the pressure at which the measurement was taken, which is why a dew point quoted without a reference pressure is only half a measurement. Field instruments often normalise to atmospheric pressure; the compartment, however, sits at several bar. A dew point that looks acceptable at 1 bar can correspond to a much higher water concentration at compartment pressure.
Practical rules that prevent most confusion:
Moisture in a sealed compartment has two sources: ingress through seals over time, and the gas handling process itself. A compartment that reads dry at commissioning and wet two years later has a seal problem. A compartment that reads wet immediately after a repair has a handling problem — most often a hose that was not purged, or a fill without evacuation.
When SF6 interrupts a fault, the arc breaks the molecule apart. Most of it recombines, but a fraction reacts with trace moisture, oxygen and contact material to form stable species that stay in the gas.
| Species | Significance |
|---|---|
| SO2 (sulphur dioxide) | The most common indicator of arcing; also forms from reaction with moisture and oxygen |
| SOF2 (thionyl fluoride) | Hydrolysis product of SF4; corrosive, and a marker of moisture involvement |
| SO2F2 (sulphuryl fluoride) | Stable end product; persists long after the fault, so it indicates historical arcing |
| H2S (hydrogen sulphide) | Indicates reaction with moisture and certain materials; strongly corrosive |
| HF (hydrogen fluoride) | Forms with moisture; attacks glass, porcelain and metal; extremely aggressive |
| Metal fluorides | Powdery deposits from contact material; usually found on internal surfaces rather than in the gas |
The interpretation is straightforward and useful. A clean compartment that has never fault-tripped shows near-zero decomposition products. A compartment that has cleared a fault shows elevated SO2 and SOF2 immediately afterwards, decaying over time as filters and surfaces absorb them, while SO2F2 persists. A compartment showing decomposition products with no record of a fault indicates either an undocumented fault trip or a partial discharge problem.
The moisture link is the important one. Decomposition products plus moisture produce acids, and acids attack exactly the surfaces that provide dielectric margin. This is why the two parameters must be read together: gas that is marginally wet is tolerable if it is clean, and gas that is clean but wet becomes a corrosion problem after the first fault.
Field instruments have improved to the point where a single analyser can report dew point, purity and SO2 in one pass. Knowing what each sensing method is good at prevents both false alarms and false confidence.
| Method | Measures | Strength | Watch out for |
|---|---|---|---|
| Chilled-mirror dew point | Moisture | Direct physical measurement, good accuracy | Sensitive to contamination on the mirror; needs clean sample |
| Capacitive moisture sensor | Moisture | Fast, portable, good for trending | Drift and saturation; needs periodic calibration |
| Electrochemical SO2 cell | SO2 | Cheap, sensitive at ppm level | Cross-sensitivity and finite cell life |
| Infrared absorption | Purity, some decomposition species | Fast, no consumables | Overlap between species; needs species-specific calibration |
| Gas chromatography | Full composition including CF4 and air | Reference-grade accuracy | Laboratory equipment; not a field method |
For routine maintenance, a portable analyser giving dew point, purity and SO2 is enough to make the release decision. For reclaimed gas destined for reuse, a laboratory analysis is the defensible route — the reuse standard covers parameters that field instruments do not measure reliably.
The single largest source of bad gas quality data is not the instrument. It is the sample.
Moisture is the parameter most easily corrupted by sampling. A hose that has absorbed water will release it into the sample and produce a wet reading on dry gas. This is why a suspiciously high dew point should always be repeated with a different hose before anyone schedules a dry-out.
| Result pattern | Likely cause | Action |
|---|---|---|
| Purity low, moisture normal, decomposition clean | Air ingress during filling or through a leak | Re-evacuate and refill; leak survey if in service |
| Moisture high, purity normal, decomposition clean | Seal ingress or wet handling | Repeat with a clean hose, then dry out or replace gas |
| Decomposition products elevated, moisture normal | Recent fault interruption | Record duty, monitor decay, replace gas if above reuse limits |
| Decomposition products and moisture both elevated | Arcing in wet gas — acid formation likely | Replace gas, inspect internals, check for corrosion |
| All parameters normal but density falling | Pure SF6 leak with no air ingress | Leak survey; the gas quality is fine, the compartment is not |
| Purity and moisture both drifting between visits | Progressive seal degradation | Plan a seal replacement before the next wet season |
The last two rows are worth remembering, because they show that gas quality and gas quantity are independent measurements. A compartment can be losing gas steadily while the remaining gas is perfectly clean — and it can hold full density while the gas inside is unfit for service.
Recovered gas follows a decision path, and it should be documented at every step.
The judgement call that causes trouble is “it’s nearly within limits, put it in the spare bay”. Reuse criteria exist because the gas will spend another twenty years inside equipment, and marginal gas degrades the margin that the breaker’s insulation design assumes.
At 40.5 kV, a compartment has generous dielectric margin and gas quality deviations are usually absorbed. At 252 kV the margin is thinner and the gas is doing more of the insulating work, which is why moisture and purity limits tighten as voltage rises.
A 252 kV pole designed for 460 kV power-frequency withstand and 1050 kV lightning impulse withstand is relying on clean, dry gas at rated density to achieve those figures. Contamination reduces the withstand level without changing the density reading, which is precisely why a density check alone is not a release criterion.
The same logic applies to interrupting capability. A 50 kA interrupter depends on gas that will quench an arc and recover dielectric strength in milliseconds; moisture and acidic decomposition products degrade both. And at the other end of the range, a 40.5 kV compartment at 0.5 MPa is just as dependent on the gas being what the nameplate says it is.
The measurement methods above are the same ones used in breaker acceptance and commissioning testing, and the acceptance criteria they serve come from the IEC 62271-100 and IEEE C37 series. Why the gas behaves the way it does — and why it decomposes under arcing at all — is covered in why SF6 gas is used in circuit breakers.
Measure three things, not one: purity, moisture and decomposition products. Read them together, because each pattern points to a different fault. Record the unit, the reference pressure and the temperature with every moisture value. Purge the sampling line and repeat any suspicious reading before acting on it. And apply the right standard — new gas, reused gas and in-service gas are three different questions.
Density tells you whether the breaker can interrupt. Gas quality tells you whether it should be allowed to try.
New gas is typically specified at 99.7% or better, with reuse criteria a point or two lower, and in-service limits set by the equipment standard. The governing figures come from the current edition of IEC 60376 for new gas, IEC 60480 for reused gas, and the breaker manual for in-service gas.
Either, provided the unit, the reference pressure and the gas temperature are recorded. Dew point is more directly meaningful for condensation risk inside a pressurised compartment; ppm is easier to trend. Never convert between them by rule of thumb.
That arcing has occurred and decomposition products are present in the gas. If there is no record of a fault trip, suspect an undocumented operation or a partial discharge problem. Combined with high moisture, it indicates acid formation and calls for gas replacement and internal inspection.
Only if it meets the reuse criteria after reclaiming and testing. Gas from a faulted compartment carries decomposition products and moisture, and must be filtered and analysed before any decision. Gas that fails the criteria goes for destruction.
Sampling variation, temperature differences, or a moisture sensor that needs calibration. Repeat the measurement with a purged line and a different hose before concluding the compartment has changed. A genuine rising trend over several visits is a seal problem.
No. Density confirms quantity, not quality. A compartment can be at full rated density and still contain air, moisture or decomposition products that reduce its dielectric and interrupting capability. Purity, moisture and decomposition product checks are part of the release decision.
