SF6 calculator: mass, moisture, leaks, recovery and gas quality
SF6 mass, dew point ↔ ppmv, leak rate against IEC 62271-1, recovery times, and acceptance criteria for new (IEC 60376) and used (IEC 60480) SF6.
SF6: mass, moisture, leaks, recovery and gas quality
Work out the kilograms of SF6 in a compartment, convert moisture between dew point and ppmv, get the leak rate and its yearly percentage, estimate recovery and vacuum times, and check whether new, used or in-service gas meets the acceptance criteria of IEC 60376 and IEC 60480.
Indicative result. The calculation is done on an Amperis server with the data in the form, which is not stored. This website may contain errors or omissions: before making any decision about an installation, the information must be checked and validated by a qualified engineer. Legal notice and terms of use
What this calculator does
Two things that are routinely got wrong in the field: how many kilograms of SF6 there really are in a compartment, and what the dew point the analyser just reported actually means.
Gas mass
For the emissions record, for planning a recovery and for deciding whether a leak matters, what you need is mass, not pressure. It is computed from the compartment volume, the pressure and the temperature by solving the van der Waals equation for SF6 (a = 7.857 L²·bar/mol², b = 0.0879 L/mol, M = 146.06 g/mol).
This is not academic refinement. SF6 departs from ideal-gas behaviour far earlier than air: at 6 bar absolute and 20 °C its compressibility factor is around 0.94, so the ideal-gas formula leaves out roughly 6% of the mass. The calculator shows that deviation so the size of the error is visible.
Temperature matters as much as pressure: the same enclosure at −10 °C and at 30 °C reads very different pressures with exactly the same amount of gas. That is why density monitors are temperature compensated, and why a “pressure drop” in winter is not always a leak.
Moisture: dew point and ppmv
After decomposition products, water is the problem with SF6: it lowers surface dielectric strength and, in the presence of arcing, forms corrosive compounds.
The calculator converts both ways using the Buck equations, over ice below 0 °C, which is how dew-point hygrometers report:
- from measured dew point to ppmv;
- from ppmv to the dew point that would be read at another pressure.
That second conversion is what prevents the commonest test error. The water content in ppmv does not change with pressure — the water molecules are what they are — but the dew point does: compressing the gas raises the vapour partial pressure and the water condenses at a higher temperature. Measuring at atmospheric pressure and comparing against a limit stated at service pressure, or the other way round, passes wet gas as good.
Leak rate
Two ways of measuring it, both in the calculator:
- Hood or enclosure (accumulation method of IEC 60068-2-17). The equipment is enclosed, the SF6 concentration in the hood is measured at the start and at the end, and L = (C1 − C0)·10⁻⁶·(Vs − V0)·p / t is applied. It comes out in Pa·m³/s and mbar·L/s; with the real gas temperature, in grams per second and per year.
- Two readings of the compartment. Pressure and temperature on two dates: the calculator solves van der Waals at each one and subtracts the masses. It is the only honest way to read a pressure drop, because comparing pressures without correcting for temperature mistakes winter cold for a leak. It also returns the pressure referred to 20 °C — what a compensated density monitor compares — and, with the alarm pressure, the years left before it trips.
With the filling mass, both give the yearly percentage, compared with the 0.5% per year that IEC 62271-1 sets as the standard value for closed pressure systems.
Recovery and vacuum times
Planning an outage means knowing how long it will take. The calculator estimates three things:
- the recovery time down to the final vacuum, in two phases: constant mass flow while the pressure is above the compressor suction pressure, and exponential decay below it with the vacuum compressor;
- the residual gas left in the compartment at that vacuum, and the percentage recovered;
- the air evacuation time down to the target vacuum with the vacuum pump, taking its ultimate pressure into account.
The flow rates of the AGRU-4X, AGRU-7, AGRU-8, AGTU and ARFU-2 come preset from their catalogues, and “Other unit” takes those of any other. The model does not include hose losses or the compressor’s falling efficiency at low pressure: real times are somewhat longer.
Acceptance criteria: new, used and in-service SF6
The fifth tab compares the analyser results with the limits that apply to each case:
| Parameter | New, pure SF6 (IEC 60376) | New, for mixtures | Used, for re-use (IEC 60480:2019) |
|---|---|---|---|
| SF6 | > 98.5% | > 99.7% | > 97% |
| Air | < 1% | < 0.2% | air + CF4 < 3% |
| CF4 | < 0.4% | < 0.08% | |
| Water | < 200 ppmv | < 200 ppmv | < 200 ppmv |
| Mineral oil | < 10 mg/kg | < 10 mg/kg | < 10 mg/kg |
| Acidity / decomposition | total acidity < 7 ppmv | total acidity < 7 ppmv | < 50 ppmv total, < 12 ppmv SO2 + SOF2 or < 25 ppmv HF |
For gas inside the equipment the IEC 60694 moisture criterion quoted in CIGRE 234 applies: dew point at filling pressure of −5 °C or lower. Moisture can be entered in ppmv, in ppmw or as a dew point at the pressure it was measured at, and the calculator does the conversion.
For used gas that fails, the tool says what to do in the IEC 60480 order of analysis: decomposition products (reclaim through filters or return), air and CF4 (not fixable on site: transfer the gas phase), moisture (dryer filter) and oil (oil filter). The 2004 edition of IEC 60480 and CIGRE guide 234 allowed 750 ppmv of water in equipment below 200 kPa absolute; the current edition has a single 200 ppmv limit.
Worked example
A 1 m³ compartment at 5 bar gauge and 20 °C, with a dew point of −40 °C measured at atmospheric pressure.
- Absolute pressure: 6.01 bar → density 38.4 kg/m³ → 38.4 kg of SF6. As an ideal gas it would have come out as 36.1 kg: 2.3 kg short.
- −40 °C at 1.013 bar is 127 ppmv.
- That same gas, at 6 bar absolute, has a dew point of −23 °C.
- Leak rate with a hood: from 0 to 10 ppmv in 24 h in a 1 m³ hood around 0.2 m³ of equipment, at 1.013 bar and 20 °C → 9.4·10⁻⁶ Pa·m³/s, about 18 g of SF6 per year: 0.05% of the 38.4 kg, well within 0.5% per year.
- Recovery with an AGRU-8 down to 20 mbar: about 1 h 36 min, with 120 g of residual gas (99.7% recovered). Then evacuating the air to 1 mbar with its 25 m³/h pump: about 17 min.
What this calculator does not do
It does not measure the gas: purity, moisture and decomposition products come from the analyser, and the calculator only compares those results with the limits of the standard. It does not replace the enclosure’s density monitor or leak location: the rate says how much is lost, not where; a detector or a camera says that. And it does not cover SF6/N2 mixtures or alternative gases, whose constants are different.
What measures it
Purity, moisture and decomposition products are measured with an SF6 gas analyser. Leak detection and the equipment to recover, store and refill the gas belong to the same switchgear testing range.
Frequently asked questions
- Why can't SF6 be treated as an ideal gas?
- Because it is a large, heavy molecule that departs appreciably from ideal behaviour already at service pressures. At 6 bar and 20 °C the compressibility factor is around 0.94: computing the mass as an ideal gas underestimates it by about 6%. On a 500-litre compartment that is nearly two kilograms missing from the emissions record.
- Why does the same gas give different dew points?
- Because the dew point depends on the partial pressure of the water vapour, and that partial pressure rises with total pressure. The same ppmv content that condenses at −40 °C at atmospheric pressure condenses above −25 °C at 6 bar absolute. That is why a dew-point figure without the pressure it was measured at means nothing.
- Is moisture measured at service pressure or at atmospheric pressure?
- It depends on the instrument and on the limit being applied. Many analysers measure at reduced pressure and then calculate; guides and limits are usually stated at a specific pressure. What matters is not to compare a measurement made at one pressure against a limit given at another — which is exactly the conversion this calculator does.
- How is the SF6 leak rate calculated with a hood?
- The equipment is enclosed in a hood or enclosure, the SF6 concentration inside is measured at the start and end of a known period, and L = (C1 − C0)·10⁻⁶·(Vs − V0)·p / t is applied, where Vs − V0 is the free volume of the hood. The temperature turns it into grams per second and per year, and the filling mass into a yearly percentage, which is what is compared with the 0.5% per year of IEC 62271-1.
- How long does it take to recover the SF6 from a compartment?
- It depends mainly on the compressor flow rate and the final vacuum. A 1 m³ compartment at 5 bar gauge (about 38 kg) with an AGRU-8 comes down to 20 mbar in just over an hour and a half according to the calculator’s model, leaving about 120 g in the compartment. Going down to 1 mbar leaves only 6 g, but lengthens the vacuum phase and needs a unit that reaches that vacuum.
- What purity must new and re-used SF6 have?
- New SF6, according to IEC 60376:2018, must contain more than 98.5% SF6, less than 1% air, less than 0.4% CF4 and less than 200 ppmv of water; for mixtures, more than 99.7%. Used SF6 can go back into equipment, according to IEC 60480:2019, with more than 97% SF6, less than 3% air and CF4 combined, less than 200 ppmv of water and decomposition products below 50 ppmv in total (12 ppmv SO2 + SOF2 or 25 ppmv HF).
- What moisture is acceptable in the SF6 of equipment in service?
- The classic criterion, from IEC 60694 and quoted in CIGRE guide 234, is a dew point of −5 °C or lower at filling pressure, equivalent to a water vapour partial pressure of 400 Pa. It leaves a margin against condensation, which starts at 0 °C (611 Pa). Because it depends on pressure, gas that complies at 2 bar may not comply at 7 bar.