Free tool · no sign-up
Shielding gas consumption calculator
Shielding gas is bought by the cylinder and spent by the second. Enter the flow rate, the arc-on time of one joint, the pre-flow and post-flow the machine is set to, and read the litres that joint costs, the cubic metres a shift costs, and how many joints one cylinder actually covers.
The interesting number is usually the share of the gas that flows with no arc burning, shown on the results card next to the joint and shift volumes. On a long seam it rounds to nothing. On tack work with ten starts a joint it can be most of the bill.
A flow rate, a stopwatch, and the valve that stays open longer than the arc
Gas-on time per joint tgas = tarc + n · (tpre + tpost), with n the arc starts
Volume per joint V = Q · tgas / 60 + Vpurge, with Q in l/min and t in seconds
Per shift Vshift = V · j / 1000 in m³, with j joints per shift
Off-arc share = (Q · n · (tpre + tpost) / 60 + Vpurge) / V · 100, in %
Cylinder content from the bottle Vcontent ≈ water capacity · filling pressure / 1000 in m³
Joints per cylinder = Vcontent · 1000 · (u / 100) / V, with u the share of the cylinder actually drawn
Cost per joint = (cylinder price + rental and delivery) ÷ joints per cylinder
Q = set flow · t = seconds · n = arc starts per joint · u = usable share of the cylinder, in % · V in litres, Vcontent in m³
There is no coefficient here and nothing to look up. The whole calculation is a flow rate multiplied by a time, and the only judgement in it is how long the valve is really open. The cylinder side is Boyle's law at constant temperature: a 50 litre bottle charged to 200 bar releases about 10 m³ at atmospheric pressure, two hundred times its water capacity. That ideal-gas route is close enough for argon and argon mixtures at ordinary cylinder pressures, and it does not describe a CO2 cylinder, which holds a liquid, is charged by mass and is quoted in kilograms.
The typical flow bands the tool checks your entry against are the ranges commonly published in equipment and consumable setting guides. They are shop guidance and carry no authority. ISO 14175 classifies shielding gases and mixtures by composition; it does not specify flow rates, and no range from any standard is reproduced on this page. On a production weld the flow rate that governs is the one on the qualified WPS.
How to calculate shielding gas consumption per joint and per cylinder
From a flowmeter setting and a stopwatch to the litres a joint costs and the joints a cylinder covers.
- Enter the flow rate the flowmeter is set to. Litres per minute, or cubic feet per hour on the imperial toggle. The band shown for the process you pick is a sanity check from equipment literature, not a specification — the qualified WPS governs the flow.
- Time the arc, then count the starts. Arc-on time is every pass of the joint added together. Arc starts are passes plus tacks plus restarts, because each one pays a pre-flow and a post-flow whether the weld lasts two seconds or two minutes.
- Add pre-flow, post-flow and any separate purge. The machine settings, not the nominal ones. Backing, trailing and chamber gas goes in the purge field in litres, since it is metered outside the torch.
- Read the joint and the shift figures. Litres per joint, cubic metres per shift, and the share of that gas that flows with no arc burning. On tack-heavy work that share is the number worth acting on.
- Turn it round with the cylinder. Enter the stated gas content, or the water capacity and filling pressure, and the tool gives joints per cylinder and shifts per cylinder. Add the invoice price and the rental to get cost per joint and cost per shift in your own currency.
- Post-flow is charged per stop, not per joint. A five-second post-flow on a two-minute pass is 4 % of the gas. The same five seconds on ten four-second tacks is more gas than the tacks themselves consume. Count the starts honestly: it is the single input that changes the answer most, and it is the one nobody records.
- Stated content beats water capacity. The volume on the delivery note is what you were sold. The litres stamped on the shoulder are the size of the bottle, and turning them into gas needs the filling pressure — a step that is right for argon and wrong for CO2, which is a liquid in there and empties at almost constant gauge pressure until it is gone.
- Usable share defaults to 100 % on purpose. No cylinder is drawn to zero: the regulator needs headroom above the set flow, and bottles get swapped before they embarrass anyone mid-shift. Left at 100 % this page reports a ceiling, and it says so on the result rather than quietly flattering the figure.
- More flow is not more shielding. Past the point where the column turns turbulent at the nozzle, extra flow entrains air and produces the porosity it was meant to prevent — see the porosity guide for what the resulting defect looks like and how it is judged. Draughts, a clogged nozzle, a leak on the suction side and a long stick-out all cost shielding without costing gas.
- An estimate is not a measurement. This page assumes the gas that left the flowmeter reached the arc. Recording the flow on every pass is what turns that assumption into evidence, and it is how a slow leak shows up as a trend rather than as a cylinder that ran out early — see welding parameter monitoring and WeldTrace, which samples gas flow alongside current, voltage and wire feed speed.
For what happens when the shield fails rather than when it is merely expensive, read the guide to porosity, its types and acceptance criteria.
Questions engineers ask about this calculation
How much shielding gas does one weld consume?
Flow rate multiplied by the time the gas valve is open, plus any separate purge. The time the valve is open is longer than the arc runs: every arc start pays one pre-flow and every stop pays one post-flow, so the gas-on time is the arc-on time plus the number of starts multiplied by the pre-flow and post-flow settings. At 15 l/min a two-minute single-pass joint with a five-second post-flow uses 15 × (120 + 5) ÷ 60, about 31 litres. Multiply by the joints in a shift for the shift total, and divide the cylinder content by the joint figure for the reverse question.
How much of the shielding gas bill is spent while no arc is burning?
On long seams almost none; on short, heavily tacked work it can be most of it. The off-arc share is the pre-flow and post-flow of every start plus any separate purge, divided by the joint total. A single two-minute pass with a five-second post-flow spends about 4 % of its gas off-arc. Ten tacks of four seconds each on the same setting spend more gas between the arcs than in them. That is why the number of arc starts is a field on this page rather than an assumption: the same flow rate and the same arc-on time can produce very different bills depending on how the work is broken up.
What shielding gas flow rate should I set on the flowmeter?
The one the qualified welding procedure specifies. This calculator shows a typical band for the process you select — the ranges commonly published in equipment and consumable setting literature — purely so an obviously wrong entry is visible, and it says so on the result. Those bands are not from a standard and they carry no authority: ISO 14175 classifies shielding gases and mixtures by composition and does not set flow rates at all. The right figure depends on nozzle diameter, joint position, stick-out, whether the work is shielded from draughts and which mixture is in the cylinder, and a correct setting can legitimately sit outside any general band.
How many joints will one cylinder of shielding gas cover?
The usable gas content divided by the consumption per joint. A cylinder holding 10 m³ is 10,000 litres of gas at atmospheric pressure, so a joint consuming 31 litres gives roughly 320 joints if the bottle is drawn completely. It never is: a cylinder is changed out with pressure still in it, and a regulator needs headroom above the set flow to hold that flow steady, so the usable share is a field here and it starts at 100 % deliberately — leaving it there gives you a ceiling rather than a plan.
Is a cylinder water capacity the same as the gas it holds?
No, and confusing the two is the most common arithmetic error in a gas budget. Water capacity is the internal volume of the bottle, stamped on the shoulder in litres — a common industrial cylinder is 50 litres. The gas content is the volume that bottle releases at atmospheric pressure, which is water capacity multiplied by the filling pressure: 50 litres at 200 bar is about 10 m³, two hundred times the water capacity. This calculator will take either, but the supplier stated content is the figure to prefer. The pressure route is Boyle law at constant temperature and is close for argon and argon mixtures at ordinary cylinder pressures; it does not describe a CO2 cylinder at all, because CO2 is stored as a liquid, charged by mass and quoted in kilograms, and its gauge pressure stays almost unchanged until the liquid is gone.
Why is the real shielding gas bill higher than the calculated consumption?
Three things this arithmetic cannot see. The surge when the solenoid opens: a pressurised hose empties into the torch far faster than the set flow, and on short welds with many starts that surge is a real share of the total. Leaks, at the hose, the fittings and the torch connection, which cost gas whenever the line is pressurised rather than only when welding. And a flowmeter graduated for one gas reading differently on another mixture, since the scale depends on the density of what passes through it. Add the gas vented at every cylinder change and the line purge after it, and measured consumption normally lands above the calculated figure rather than below it.
Does raising the shielding gas flow rate stop porosity?
Only when the cause is genuinely too little gas, and beyond a point more flow makes it worse. Too high a flow becomes turbulent as it leaves the nozzle and draws air into the shield, so the failure looks exactly like the shortage it was meant to fix. Porosity that persists at a sensible flow is usually pointing somewhere else: a draught across the cell, a spatter-clogged nozzle, a leaking hose drawing air in on the suction side, moisture or coating on the joint, or too long a stick-out. Turning the flowmeter up is the cheapest thing to try and the most expensive habit to keep, because it raises the gas bill on every weld regardless of whether it fixed anything.
The flow you set is not always the flow the arc gets.
A flowmeter setting is an instruction. WeldTrace records shielding gas flow at 10 Hz alongside current, voltage and wire feed speed on every pass, so a hose that started leaking on Tuesday is a measurement from the seam rather than a cylinder that emptied a week early for no reason anybody wrote down.
