Module 03 · Abnormal Operations
Flame-out Response
Procedures for responding to loss of flame — from a single burner trip to a full firebox blackout. The correct response depends on the scope of the flame loss and whether the BMS has already acted. Always establish what you actually have before touching any fuel valve.
Danger — unburnt fuel in a hot firebox
A flame-out with fuel gas still flowing is a pre-explosion condition. Your first action is always to remove the fuel, not to re-ignite. Confirm firebox negative draft before approaching any opening. Do not re-enter the firebox or attempt re-ignition until the full purge sequence is complete and all re-ignition gate checks are satisfied.
Select scenario
The response procedure differs significantly depending on the scope of the flame loss. Select the scenario that matches what you are seeing.
Single burner flame loss
One burner has lost flame. The remaining burners are lit and the firebox temperature is stable. The BMS may or may not have tripped that burner's SSOV depending on the flame scanner state.
Check the BMS first. If the BMS has registered a flame failure on that burner, its SSOV will have closed automatically. Confirm on the BMS panel before touching manual valves. If the BMS shows no trip, the fuel may still be flowing — close the manual valve at the burner immediately and treat this as a manual flame-out.
?
Confirm flame is actually out
Check via peephole before any action. A BMS scanner fault can signal flame loss with the burner still lit. If you can see flame through the peephole, treat as scanner fault — do not close fuel valve based on a BMS alarm alone.
1
Close manual fuel valve at that burner
Immediately isolate fuel to the affected burner. Do not attempt re-ignition first. If SSOV has already closed, confirm the manual valve is also closed as a secondary barrier.
2
Notify shift supervisor — log time and burner ID
Record: burner number, time of flame loss, process conditions at time of event (COT, load, fuel pressure), and BMS status.
3
Investigate cause before re-lighting
Do not re-light until cause is known. Common causes: low fuel pressure at that burner; blocked tip; air register too open (lean lift-off); excessive draft; pilot failure. See cause table below.
4
Re-light per site ignition procedure — with adjacent burner lit as witness
For a single burner in a lit firebox, full re-purge is not required. Confirm: fuel valve closed at burner, local area clear, pilot established, then open manual valve per site procedure. Monitor flame stability for ≥5 minutes after re-light.
Single Burner Flame Loss — Causes and Responses
| What you observe | Likely cause | Response before re-light |
|---|---|---|
| Flame and pilot both out; fuel pressure low at that burner | Low fuel pressure — supply issue or partial SSOV closure | Check fuel header pressure. Confirm SSOV fully open. Do not re-light until pressure is within normal range. |
| Flame out; pilot lit; main fuel valve open — no fuel pressure issue | Blocked or worn tip; fuel gas quality change (high inerts) | Close manual valve. Inspect tip. Replace if blocked or damaged. Check fuel gas quality if multiple burners affected. |
| Flame unstable / pulsing then extinguished; no fuel pressure issue | Air register too open; excessive draft at that burner location | Partially close air register before re-light. Check draft reading — reduce damper if excessive. |
| BMS trips fuel; scanner alarm; no visible flame loss on peephole inspection | Scanner fouling, misalignment, or cable fault | Verify flame is present. If flame confirmed, raise instrument work order. Do not bypass flame scanner without engineering authorisation and PTW. |
| Flame out after fuel gas composition or pressure surge event | Sudden fuel gas composition change (slug of liquid, high inerts) | Check fuel gas knock-out drum level. Drain if required. Notify process engineer if repeated. |
Partial flame loss — multiple burners out
Two or more burners have lost flame, or an entire row/pass of burners is out. The firebox is still hot and some burners remain lit. This is a higher-risk condition than single burner loss because unburnt fuel from the affected burners can accumulate faster in a still-hot firebox.
Warning — partial flame loss escalates quickly
A partial flame loss can transition to a full blackout if the cause is a system-level fuel, air, or draft issue. Treat the root cause, not just the symptom. If more burners are going out while you're responding, initiate the Total Flame-out procedure immediately.
1
Close manual fuel valves on all affected burners immediately
Do not attempt to re-light while other burners are extinguishing. Isolate all burners with confirmed flame loss. If the BMS has tripped a group SSOV, confirm it on the BMS panel.
2
Assess remaining flame — is the firebox still lit?
If any burners remain lit and the firebox is stable, you are still in partial loss scenario. If the firebox has gone dark — switch immediately to Total Flame-out procedure.
3
Reduce load — lower COT setpoint or cut feed if required
With fewer burners, heat input is reduced. Protect process tubes by reducing load to match remaining firing capacity. COT dropping rapidly is acceptable; COT rising with fewer burners on is not — investigate immediately.
4
Identify and correct root cause
Check: fuel header pressure (low pressure = likely cause of multiple burner loss); draft (excessive draft causes lean lift-off across multiple burners); fuel gas composition; ambient conditions (sudden wind increase can destabilise all floor-fired burners simultaneously).
5
Re-light affected burners one at a time — from lit pilots
Re-light into the lit firebox. Start with the burners closest to operating burners. Confirm pilot on each burner before opening main fuel valve. Re-light in sequence — do not open multiple burner fuel valves simultaneously.
Total flame-out — firebox blackout
All flames are out. The firebox is dark. This is the highest consequence flame-out scenario. A hot firebox with no flame and fuel gas present is a pre-explosion hazard. The BMS should have closed the main fuel trip valve (MFTV) and all SSOVs — verify this immediately.
Danger — do not attempt to re-light without completing full purge
Attempting to re-ignite a firebox that has not been purged of accumulated gas can cause a firebox explosion. The purge requirement is not waivable. Minimum purge: 5 × firebox volume air changes at the site-specified purge rate — typically 5–15 minutes depending on heater size. Follow site-specific BMS logic for purge confirmation.
Total Flame-out — Immediate Response Sequence
01
Confirm MFTV closed and all SSOVs closed — BMS panel check
Verify the BMS has tripped on flame failure. If any SSOV remains open, close it manually. Do not proceed until all fuel to the firebox is confirmed isolated. If BMS has not tripped, initiate manual trip from the BMS panel.
02
Notify shift supervisor and control room immediately
State: "Firebox blackout confirmed on [Heater tag]. Fuel isolated. Beginning purge sequence." Time-stamp in the shift log. Do not delay notification.
03
Protect process — reduce feed or maintain minimum flow
Do not allow process fluid to stagnate in tubes. Maintain minimum flow through all passes if safe to do so. If no flow can be maintained, initiate emergency depressure or ESD per site procedure. A dead tube in a still-hot firebox is a coke-out or rupture risk.
04
Confirm negative draft — do not open any firebox access
Check draft gauge. Firebox must be at
negative pressure before approaching any peephole, burner front, or inspection door. If draft is positive or unknown, do not approach the firebox.
05
Initiate purge — confirm purge conditions met
Open stack damper fully. Confirm forced-draft or natural-draft airflow through firebox. Start purge timer per BMS logic. Minimum purge volume is
5 × firebox volume at verified air flow rate. Do not proceed to re-ignition before BMS purge permissive is satisfied.
06
Investigate cause — do not re-light without understanding why the blackout occurred
Likely causes: total fuel gas failure (header trip, ESD from upstream); BMS spurious trip; loss of all pilots in severe wind event; total loss of combustion air (forced-draft fan trip). If cause is unknown, do not re-light until root cause is identified.
07
Complete re-ignition gate checks — then restart per startup procedure
All gate checks below must pass before the BMS will allow re-ignition. After gate checks complete, follow the full startup procedure (fh-startup.html) — not a shortcut sequence.
The purge requirement
Purge is not a formality. A typical firebox explosion involves a fuel–air mixture in the flammable range (approximately 5–15% v/v for natural gas) ignited by re-ignition sparks or hot refractory. The purge achieves two things: it dilutes any accumulated fuel gas below the lower flammable limit, and it carries un-ignited gas out of the firebox before a source of ignition is introduced.
5–15
minutes typical
Minimum 5 × firebox air changes at confirmed flow rate
BMS purge timer must complete — not manually overrideable without PTW
Stack damper must be fully open during purge
All fuel valves (MFTV, SSOVs, manual burner valves) must remain closed throughout
Larger fireboxes require longer purge — confirm against site-specific BMS logic
Re-ignition gate checks
All items below must be confirmed before the BMS re-ignition permissive is available. This is an interactive checklist — use it in the field to confirm readiness.
Re-ignition Gate — All Must Be Confirmed Before Proceeding
Post flame-out — what happens to the process
A flame-out — even one quickly resolved — affects the process and the heater. These consequences must be managed regardless of how fast the re-light is achieved.
Process and Equipment Consequences — Flame-out Events
| Parameter | What happens during flame-out | Operator action |
|---|---|---|
| COT (Coil Outlet Temp) | Falls as heat input is lost. Rate of drop depends on heater thermal mass and feed rate. | Monitor rate of COT drop. If COT approaches process minimum, notify downstream unit and consider reducing feed. Do not raise fuel flow ahead of confirmed flame to recover COT. |
| Process pressure | May rise if downstream restriction reacts to lower temperature / flow changes. | Monitor pass inlet and outlet pressures. Ensure relief paths are available. |
| Tube skin temperatures | Fall initially as heat input is lost. Hot refractory can continue to radiate heat to tubes for 15–30 min after blackout. | Do not assume tubes are safe simply because burners are out. Monitor tube skin temps where available during purge period. |
| Firebox / refractory temperature | Begins to fall once firing is lost. Refractory retains heat for an extended period. A re-light into a hot firebox restores heat faster than a cold start. | Note time of blackout. Communicate firebox temperature status (hot/warm/cold) to shift supervisor and anyone assisting re-start. |
| BMS state | Following a flame failure trip, the BMS enters a lockout state. It will not allow re-ignition until the purge sequence has been completed and all permissives are met. | Follow BMS reset procedure as specified in site documentation. Do not attempt to bypass or defeat BMS logic to accelerate restart. |
Reporting requirement — all flame-out events
Every flame-out — including single burner — must be recorded in the shift log with: time of event, heater tag, scope (single/partial/total), BMS response, root cause assessment, corrective action taken, and time of successful re-light. Total blackouts typically require a formal incident report in addition to the shift log entry.