When the Duct Burner Flame Touches the Tubes

Oct 5

A duct burner flame is supposed to finish burning before it reaches anything. Here is why it sometimes does not, what that costs, and how the fix is engineered.

A flame with nowhere to hide

In a fired heater, the flame gets a firebox: a tall radiant section designed around it, with the tubes kept at a distance. In an HRSG, the duct burner flame gets a duct. It burns crosswise in a fast-moving river of turbine exhaust, and a few feet downstream stand the superheater tube banks it exists to serve. The design assumes combustion completes in the space available. When it does not, flame and unburned combustion wash directly against tubes and liners, and the hottest components in the plant run hotter than anyone calculated.

Readers of our duct burner introduction will remember the basic bargain: the burner borrows its oxygen from the turbine exhaust and its mixing energy from the exhaust flow. Impingement is what happens when that bargain breaks down locally.

Why it happens

Uneven inlet flow. A duct burner is designed for a uniform velocity profile across its face. What it receives depends on the duct upstream: the turbine exhaust expands through a transition, turns, and arrives with fast and slow regions. Burner vendors are blunt about this. Good flow distribution is required to keep the flame stable and short and to protect side-wall liners and downstream components from impingement, which is why distribution grids and baffles sit upstream of the burner and why that grid is on the standard inspection list. Where the approach flow is slow, the flame lengthens; where it is skewed, the flame leans.

Not enough mixing length. Flame length is set by how fast fuel finds oxygen. Low local O2 (a heavily fired unit, or one behind an efficient turbine), cold spots, or a damaged element all stretch the flame. If the stretched flame is longer than the distance to the first tube bank, the tubes become the flame holder.

The uprate multiplier. As we covered in the GT upgrades article, an uprated turbine sends more exhaust, hotter, through the same duct. The burner now fires into a faster stream with a different oxygen margin and a distorted velocity profile it was never tuned for. A burner that behaved for fifteen years can start misbehaving the month after an uprate, and nobody changed the burner.

What it costs

Three kinds of damage follow, and the most expensive one is the hardest to see.

Local overheating. Impingement is a local heat flux spike, and tube metal temperature follows flux directly. Using the resistance-ladder arithmetic from our tube metal temperature article: raise the local flux by half and the temperature rise across every layer scales with it, so a tube designed with a modest margin spends its hottest hours well past its design metal temperature. Creep life trades exponentially against temperature, and a few tens of degrees of sustained overtemperature consume life at several times the design rate. Combined Cycle Journal has warned about this pattern in duct-fired HRSGs in an article titled "Beware hidden superheater, reheater overheating": the tubes look fine, the steam temperatures look fine, and the damage accumulates where nobody has a thermocouple.

Liner and casing damage. A flame leaning into a side wall cooks the liner and its insulation. This is the visible failure, the one found at inspection, and it doubles as the diagnostic: scorched liner panels downstream of one burner element are a map of where the flame has been going.

Steam temperature control problems. A distorted flame heats the tube bank unevenly, so parallel tube circuits pick up different duties. The attemperator then works harder to hold the header temperature, which causes tube damage of its own, and individual hot circuits can run past their design point while the mixed steam temperature reads normal. That is the "hidden" in hidden overheating.

Seeing scorched liners downstream of your duct burner, uneven tube-row temperatures, or a flame that got longer after a turbine uprate? Send us photos and your operating data, and an FIS engineer will give you an honest read on what is happening. Tell us about your unit!

The fix is engineering, not luck

This problem is getting attention. At POWERGEN International 2025, a project team presented a full duct burner replacement on an operating HRSG, aimed at reducing repair costs and flame impingement at the same heat input, with inspection results after 16 months of operation. Whether the answer on a given unit is replacement, modification, or tuning, the engineering sequence is the same.

Model the gas path. CFD of the inlet duct, distribution grid, and burner plane shows the velocity and oxygen profile the burner sees, for the real turbine exhaust including any uprate cases. This is the step that explains the scorch marks.

Fix the flow before blaming the burner. Distribution grids and baffles sized from the model flatten the profile. Many burner problems turn out to be duct problems.

Match the burner to the duct. Element type, spacing, and firing pattern set the flame envelope. The flame has to be finished before the first tube row, and the burner selection has to prove that at the worst-case operating point rather than the average one.

Verify the oxygen margin. Firing duty and the exhaust's remaining O2 set how hard each element works. The oxygen check from our duct burner introduction is the first screen; a unit firing near its oxygen limit runs long, lazy flames.

Set firing limits against tube metal, not steam. The last defense is operational: a firing-rate limit derived from the calculated tube metal temperature at the hottest row, not from the steam header thermometer that averages the problem away.

We have been fixing this flame for 30 years

Flame impingement on process tubes is one of the oldest failure modes in fired heater practice, and it is the problem our patented Inclined Firing System exists to solve. On one vacuum heater, flames from vertically fired burners were rolling onto the tubes. Inclining the firing geometry ended the impingement, and the revamped heater went on to gain 16% capacity with tube metal temperatures held within limits. The case study is here. In an HRSG duct the geometry is different but the physics is the same: a flame shaped by the flow and the oxygen field, and a boundary it must not cross. Engineering the flame envelope is the job on either machine.

The bottom line

A duct burner flame that reaches the tubes has a cause: flow distribution, mixing length, or an operating point the burner was never designed for. Each one is calculable and fixable. The expensive version is the one where nobody looks until a tube fails, because creep damage accumulates while every gauge reads normal. If your liners are scorched or your flame changed after an uprate, it is worth a look now.

FIS provides independent engineering for fired heaters and HRSGs, including CFD of inlet ducts and burner planes, flow distribution design, burner evaluation, and tube metal temperature rating. Contact us to scope a duct burner review.

vimla pawar