A burner hose almost never fails on pressure. A piped natural gas train runs at 0.1–2 bar and a furnace oil line at 5–10 bar, against a DN25 assembly rated 38 bar cold and still good for 23 bar at 200 °C. The pressure has three or four times the margin it needs. What kills burner hoses is heat conducted back along the fitting, the swing of the burner on its hinge, and a joint that was crimped rather than welded.
What this service does to a hose
The first mechanism is heat arriving where nobody expected it. The fuel itself is cold or warm, but the burner body is not, and the last two hundred millimetres of hose sits in radiated heat from the furnace face and conducted heat from the burner casting. The ferrule is the thickest part of the assembly and the least able to shed that heat. Assemblies that were sized on the fuel temperature and failed at the collar were usually sized on the wrong temperature entirely — the number that matters is the metal temperature at the fitting, not the temperature of what is inside it.
The second is movement, and it is the reason a hose is there at all. Most burners are hinged so they can be swung clear for tube cleaning or refractory work. Every swing is one full cycle of the hose at whatever bend radius the installation allows. Do that twice a shift for five years and the assembly has seen several thousand cycles — not many, unless the hose was cut short and is being worked below its flexing bend radius, in which case a few thousand is plenty to crack a convolution root.
The third is the end joint. A crimped or swaged collar relies on mechanical grip, and mechanical grip relaxes when it is thermally cycled. On a fuel line the joint is part of the pressure boundary and it is carrying something that burns. We weld ours for that reason and for no other.
What we would fit
| Element | What we would supply | Why |
|---|---|---|
| Hose | 316L annular corrugated. 321 where the burner front runs hot and cycles daily | 321 is stabilised, so long periods in the sensitisation range do not cost it corrosion resistance |
| Braid | 304 wire, single braid on gas; double where the line is also a vibration path | Single braid clears every burner pressure we have seen by a wide margin |
| Ends | Male hex nipple one end, swivel flange PN16 the other, as standard | The swivel lets the flange be lined up without twisting the hose, which is the commonest installation fault |
| Attachment | TIG welded, never crimped | The joint holds fuel and it is thermally cycled |
| Length | 1000, 1250 or 1500 mm standard, sized on the swing of the burner | The hose has to stay above its flexing radius at the extreme of travel, not at rest |
What we would not fit, and what happens
Interlock hose. The seam is mechanical, not welded. It is not pressure-tight and never was. On a fuel line that is not a performance question, it is a safety one. See interlock construction.
A hose cut to the shortest length that reaches. The classic burner failure. It looks tidy at commissioning and it is being worked below its bend radius the first time the burner is swung out.
A hose installed in line with the swing. A hose takes bending and it takes offset. It does not take torsion, and swinging a burner about a hinge puts torsion into anything mounted along the hinge axis. Install across the movement — the movement types page shows the geometries that survive.
Anything we have not been told the fitting temperature for. If the enquiry says “natural gas, ambient” we will ask again, because the gas temperature is not the hose temperature on this duty.
What we need to size it
| We need | Because |
|---|---|
| Fuel — PNG, CNG, LPG vapour, biogas, furnace oil, LDO or HSD | It decides the grade and whether a liner is worth having |
| Line pressure, and the relief or regulator setting | The relief setting is the real design pressure, not the running one |
| Maximum temperature at the hose, not at the burner | This is the number the derating runs on and the one most often missing |
| How far the burner swings, and how often | It sets the length and whether the flexing or the static bend radius governs |
| Thread form and size at the nipple end, flange standard at the other | BSP, BSPT and NPT are not interchangeable and the flange may not be PN16 |
| Whether test certificates travel with the order | It changes the documentation route, not the assembly |