03 · Abnormal Operations
Tube Failure / Leak
Recognition, classification, and response to process tube failures — from early-warning seeps to catastrophic rupture. The speed and nature of your response depends entirely on correct identification.
Any confirmed tube leak = immediate heater trip consideration
A process tube failure releases flammable hydrocarbon into an active firebox. Ignition can occur instantly. Do not attempt to manage a confirmed leak while the heater remains in service unless your site procedure explicitly permits a controlled run-to-trip with active monitoring. When in doubt, shut it down.
Why Tubes Fail
Tube failures are rarely sudden without precursor. Most develop over operating cycles from one or more interacting mechanisms. Recognising the mechanism helps predict where the next failure will occur and what the leak profile will look like.
Thermal
Overheating / Creep
Sustained metal temperature above design limit causes creep — slow plastic deformation under stress. Tube bulges, then ruptures. Common cause: loss of flow, maldistribution, or coke hot spots. Often shows bright glowing tube before failure.
Mechanical
External Corrosion / Oxidation
High-temperature oxidation thins the tube OD. Sulfidation attack in high-H₂S environments. Accelerated by wet cycles and insulation faults at return bends. Failure is often localised at bends or hangers.
Process-Side
Internal Corrosion / Erosion
Acid corrosion, naphthenic acid attack (high-TAN crudes), or erosion from solids/velocities thin the tube ID. Wall loss is insidious — no external sign until a pinhole leak develops. Detected only by inspection or sudden failure.
Cyclic
Thermal Fatigue
Repeated heat-up/cool-down cycles produce differential expansion stresses, particularly at welded connections, return bends, and supports. Initiates cracks rather than wall thinning. Units with frequent startups are most at risk.
Deposition
Coke Localisation
Localised coke build-up insulates the tube from the process flow, causing the metal temperature to rise sharply under that patch. A bright hot spot visible through a peephole is a coked tube section trending toward failure.
Mechanical
Flame Impingement
Direct burner flame contact on a tube surface produces localised overheating far above bulk process temperature. Results in rapid creep failure at a discrete, well-defined location. Usually accompanied by a visible bright spot.
Recognition — Signs by Severity
Tube failures present across a wide spectrum. Knowing which signs correspond to which stage of failure determines whether you have minutes or seconds to respond.
Tube Failure Signs — Field Recognition Guide
| Sign | What it means | Urgency |
|---|---|---|
| Unexplained rise in COT (coil outlet temperature) at constant duty | Possible loss of flow in one pass — stagnation and overheating developing | Investigate now |
| Bright glowing spot or bulge visible through peephole | Metal temperature significantly above normal — creep or coke hotspot. Pre-failure condition. | Imminent failure |
| Sudden drop in pass outlet temperature or flow | Tube rupture in that pass — loss of process fluid into firebox | Failure occurring |
| Abnormal flame colour — sudden change to luminous yellow or orange | Process hydrocarbons entering the firebox and burning | Failure occurring |
| Black or dense smoke from stack | Rich combustion from excess hydrocarbon — process leak burning in firebox | Failure occurring |
| Firebox pressure positive or unstable draft | Large hydrocarbon release disrupting combustion air balance | Major failure |
| Loud bang or pressure surge felt/heard at heater | Catastrophic rupture or ignition of accumulated vapour | Evacuate immediately |
| Fire visible outside casing, at stack base, or around peepholes | External fire from escaping process fluid | EMERGENCY — activate ESD |
| Hydrocarbon smell near heater at grade without visible flames | Small seep or pinhole from return bend, fitting, or weld — possible slow internal leak to atmosphere | Urgent investigation |
| Instrument freeze — COT or pass flow reading locks or goes to fail-safe | Tube failure may have damaged thermowell or instrument — do not dismiss as instrument fault without investigation | Investigate — do not assume instrument |
Never approach peepholes head-on during a suspected tube leak
Approach from the side. If firebox pressure has gone positive, opening a peephole can cause flame or hot gas ejection. Confirm draft is negative before any peephole inspection.
Response Procedures
Severity: Medium
Abnormal signs present but no confirmed process release into firebox
Suspected Tube Leak Response
01
Notify control room and supervisor immediately
Do not investigate alone. Advise what you observed and from where. The decision to continue investigation or initiate shutdown must involve supervision.
02
Review DCS trends — all pass flows, COT, TMT readings
Look for: one pass diverging from others, unexplained COT rise at constant fuel, or a single thermocouple reading high. A maldistributed or blocked pass will show reduced flow and elevated temperature.
03
Approach from upwind side; check for hydrocarbon odour at grade
Before reaching the heater, stop and assess wind direction. Any hydrocarbon smell in the approach path is significant. Do not enter the hazard area without a gas detector and a second person.
04
Peephole inspection — side approach only, confirm negative draft first
Check draft gauge at peephole before opening.
Draft must be negative (typically −1 to −5 mmWC) before opening. Look for abnormal flame colour, any tube showing brighter than surroundings, or visible bulge. Do not linger.
05
Check casing skin temperature for hot spots
An infrared gun or hand-check of external casing. A hot spot on the casing (unable to hold hand on) can indicate refractory failure or a very localised tube overtemperature behind that panel.
06
Reduce heater duty to minimum stable firing
Lower fuel gas setpoint — do not extinguish burners, but reduce heat input. This buys time for investigation and reduces severity if a leak develops. Monitor COT — if target cannot be maintained at reduced duty, suspect loss of flow in a pass.
07
Decision point: confirm or rule out leak
If DCS data, visual inspection, and gas detector are all normal and no source identified — document and continue enhanced monitoring. If any indicator confirms process release, move immediately to Confirmed Leak procedure. Do not delay this decision.
08
Log all observations with timestamp
Record exactly what was observed, from where, at what time, and what the DCS values were. This is essential for post-incident review regardless of outcome.
Severity: High
Process fluid confirmed releasing into active firebox — controlled emergency shutdown required
Time-critical — do not delay shutdown to diagnose cause
Once a leak into an active firebox is confirmed, the priority is isolating fuel, depressurising the process, and smothering the firebox. Diagnosis happens after the event is controlled.
Confirmed Tube Leak — Emergency Shutdown
01
Initiate emergency shutdown — trip all burners
Activate BMS emergency shutdown or manually close main fuel gas block valve. Every second of continued firing increases heat input to a failing tube. Do not reduce gradually — trip immediately.
02
Activate firewater / steam smothering to firebox if available
Some heaters have steam injection to the firebox or convection section to suppress remaining combustion and cool the box. Activate per site procedure. Do not inject if it could cause a steam explosion against a very hot tube.
03
Isolate process feed to heater — close inlet isolation valve(s)
Close the heater charge isolation valve. This stops fresh hydrocarbon entering the leaking coil under pump pressure. Confirm on DCS that flow drops to zero. If valve fails to close, advise supervisor — alternate isolation may be required upstream.
04
Depressurise and deprime the coil
Open the coil blowdown valve to flare. Maintain minimum residual flow if possible — if the coil is still at temperature, stagnant hydrocarbon will coke and block the tube, worsening the failure. Some procedures call for a nitrogen purge to displace remaining hydrocarbons.
05
Close stack damper to limit air ingress to firebox
Once burners are out, closing the stack damper (partially or fully) limits oxygen reaching any residual hydrocarbons in the firebox, reducing fire risk. Do not fully close if the coil is still hot — natural convection is needed for cooling. Follow site-specific procedure.
06
Account for all personnel — headcount and muster
Confirm all operators in the area are accounted for before any further field action near the heater. No one approaches the heater until confirmed safe by supervision and emergency services (if called).
07
Notify plant emergency response team and shift superintendent
Escalate per site emergency notification procedure. Provide: heater tag, confirmed or suspected cause, current status of fuel/process isolation, and whether there is an external fire.
08
Monitor firebox temperature — do not allow cooldown to be too rapid
Rapid cooldown of a hot refractory can cause cracking. If the box is cooling naturally (no external fire), allow a gradual cool. Monitor skin temperatures. Do not introduce cooling water into the firebox unless specified — thermal shock will damage refractory lining.
09
Verify all fuel sources isolated — gas test the area
Once the box has cooled sufficiently, perform a gas test at grade around the heater perimeter, at stack base, and at any drains or sumps. Do not declare the area safe until reading is below LEL.
10
Preserve evidence — do not alter any settings before incident review
Leave all valve positions, damper settings, and DCS states as-found until the incident investigation team has reviewed them. Capture screenshots of trend data. Note exact time of all actions taken.
Severity: Critical
Catastrophic failure or external fire — life safety priority, evacuate and activate emergency response
Evacuate first — operate second
Do not attempt to manually operate valves or equipment while an external fire is active or a rupture is in progress. Activate ESD from a safe distance or the control room. Personnel safety takes absolute priority over equipment preservation.
Rupture / External Fire Response
01
Raise the alarm — activate site fire alarm and evacuate the area
Sound the nearest manual call point. Evacuate all personnel upwind to the muster point. Do not re-enter the exclusion zone. Activation of the general site alarm triggers emergency response team mobilisation.
02
Activate Emergency Shutdown (ESD) from control room or remote ESD panel
Initiate full ESD from a safe location. This simultaneously trips fuel gas, isolates process feed, and activates automatic fire suppression if installed. Do not attempt manual valve operation in the fire zone.
03
Do not attempt to fight the fire if it is fed by process flow
A jet or pool fire from a pressurised leak must not be extinguished unless the source is simultaneously isolated. Extinguishing a gas jet fire without isolating the source creates an unburned vapour cloud — which can re-ignite with far greater force (BLEVE / VCE risk).
04
Apply cooling water to adjacent vessels and structures — not directly into the fire
Protect adjacent equipment from radiant heat using fixed deluge or manual hose streams directed at the exposed surfaces, not at the base of the fire. Prevents domino failures and BLEVE of adjacent pressure vessels.
05
Account for all personnel — report to incident commander
Headcount at muster point. Report to the incident commander any personnel who are unaccounted for. Do not re-enter for any reason until the incident commander authorises.
06
Provide process information to emergency services
The incident commander or shift superintendent provides: process fluid type and approximate pressure/temperature at failure, isolation status of fuel and process, any secondary hazards (H₂S, HF, steam), and heater dimensions for exclusion zone assessment.
07
Do not restart or enter heater until formal investigation clearance is given
Following any fire or major tube rupture, a formal incident investigation and mechanical integrity inspection is required before any heater entry or restart. This is a management of change (MOC) event — no exceptions.
Quick Decision — Which Response?
Use this when you first arrive at the heater with an abnormal condition.
?
Is there a fire external to the heater casing, or a loud bang/rupture?
YES →Rupture / External Fire procedure. Evacuate first.
?
Can you see black smoke from stack, or abnormal flame visible through peephole (yellow/luminous), or significant gas detector reading at grade?
YES →Confirmed Leak procedure. Trip the heater immediately.
?
DCS showing one pass diverging, unexplained COT rise, hydrocarbon smell, or a glowing hotspot through peephole?
YES →Suspected Leak procedure. Reduce duty, investigate, supervise.
✓
None of the above — all indicators normal
Continue enhanced monitoring. Document what triggered the inspection. Review TMT and pass flow trends over the last shift. If anything changes, re-enter this tree.
After the Event — Before Restart
No fired heater that has experienced a tube failure may restart without completing the following minimum checks. Site management and inspection authority sign-off is required.
Minimum Pre-Restart Checks After Tube Failure
| Check | Method | Responsibility |
|---|---|---|
| Visual inspection of all visible tube sections — bulge, discolouration, sag | Heater entry (cold box, gas-free) with inspection lighting | Operations + Inspection |
| Tube wall thickness measurement at and around the failure point | UT or RT by certified inspector | Inspection (NDT) |
| Adjacent tube assessment — at minimum ±2 tubes each side of failure | UT, visual, and where possible strain measurement | Inspection (NDT) |
| Refractory inspection — cracking, spall, fallen sections | Visual from firebox entry | Operations + Maintenance |
| Return bend and fitting inspection | Visual and PT/MT for crack detection at welds | Inspection |
| Burner tip and tile inspection | Visual — look for coke build-up or damage from the event | Maintenance / Operations |
| Cause determination documented and corrective action defined | Incident investigation report | Engineering + Operations Management |
| MOC (Management of Change) raised and approved for restart | Site MOC procedure | Engineering / Plant Manager |
Root cause, not just repair
Replacing a failed tube section without understanding why it failed almost always leads to a repeat event. The investigation must establish whether the failure was thermal (duty, flow, coke), metallurgical (material selection, corrosion, fatigue), or operational (startup rate, flame impingement). The answer determines what changes before restart.