Crude Preheat Heater Tube Rupture: Characterizing Overheating Before Return to Service
A 5 in OD × 0.250 in minimum-wall carbon steel tube in a crude unit preheat heater split longitudinally after 11 years of service. The rupture is roughly 30 in long and sits in a visibly bulged section; the fracture face is thick, heavily oxidized, and rough, with a short shear lip at each end. The tube has grown from 5.000 in to 5.230 in OD at the failure plane. Design tube-metal temperature was 900°F. Pass 3 of the heater has a documented history of flame impingement, and an optical pyrometer survey three months earlier read 1080°F on this exact tube row. Metallography reports advanced pearlite spheroidization and carbide coalescence near the fracture, with no evidence of sigma phase. The area operations manager wants the failed tube cut out, the spare 12 ft section welded in, and the heater returned to service within 48 hours.
1. Classify the failure mode and explain what the fracture appearance, dimensional data, and microstructure each rule in or out. What does the design tube-metal temperature tell you?
Read the worked answer
This is a long-term (creep) stress-rupture failure, not short-term overheating.
Discriminating evidence:
- Thick, oxidized, rough fracture face with a small end shear lip - the crack grew slowly over months to years by creep, then the remaining ligament overloaded. Short-term overheating produces a thin, knife-edge rupture with essentially no bulging because there is no time for creep strain to accumulate.
- 4.6% OD growth (5.000 → 5.230 in) is well beyond the ~1–2% creep strain that already signals significant life expenditure; short-term overheating typically shows little or no diametral growth.
- Pearlite spheroidization and carbide coalescence require sustained time at temperature (roughly >950°F for this steel), confirming chronic exposure rather than a brief excursion.
- Absence of sigma phase is expected for carbon steel and does not argue against creep.
The design 900°F tube-metal temperature is itself the root of the problem. Carbon steel begins to creep meaningfully above about 800°F, so the tube was designed with essentially no creep margin and only survived 11 years because actual metal temperatures were mostly lower than design. Sustained operation at 1080°F from flame impingement cut the life short. The mechanism is correctable only by fixing the thermal condition, not by welding in a new tube.
2. Replacing the single failed tube will restore pressure integrity but not reliability. What extent-of-condition work must be completed before the unit restarts, and how should remaining life of the other tubes in the pass be judged?
Read the worked answer
The failed tube is one sample of a population that has all seen the same thermal history. Before restart:
- Dimensional survey of the whole affected pass (and adjacent passes). Measure OD at a defined grid - every tube, several axial locations, including the highest-flux zone. Compare against the original and against API 530/API 579 creep-remaining-life criteria. Tubes exceeding the strain criterion (commonly ~2% or per the applicable fitness-for-service assessment) should be replaced, not monitored.
- Tube-metal temperature mapping. Repeat the pyrometer/IR survey under steady full-rate firing and correlate to burner position; identify every impinged tube, not just the failed row. Skin thermocouples on the worst tubes give continuous confirmation.
- Metallurgical sampling. Field replication or tube sections from high-temperature zones upstream and downstream of the failure, to grade spheroidization and check for graphitization and internal oxide scale thickness - scale growth is a time-at-temperature indicator.
- Reassess the whole coil against the original design basis. With design TMT at 900°F, the remaining tubes are already outside a defensible creep envelope. A 48-hour weld-in of one section restores containment while leaving the population unqualified; the honest answer is that the pass needs replacement planning concurrent with burner correction.
3. What operating and design changes would you specify so that a repaired coil does not repeat the failure, and what monitoring would you put in place to detect recurrence early?
Read the worked answer
The creep damage is thermally driven, so mitigation is thermal:
- Eliminate flame impingement - burner alignment, tip replacement, flame pattern survey, and possibly a burner upgrade or altered firing arrangement for pass 3. Verify by visual flame scan plus post-change TMT survey.
- Reduce peak flux and TMT - derate firing or redistribute duty so measured TMT has margin below the creep regime; an alarm/trip setpoint on tube skin temperature (e.g., consistent with design plus a defined margin) plus a low-flow trip on the pass.
- Upgrade material where temperature cannot be controlled - a low-alloy or higher-Cr tube alloy provides creep strength the carbon steel never had at 900°F design.
- Monitoring: permanent skin thermocouples on the hottest tubes with trending and alarms; periodic IR scans at defined intervals; periodic OD/diametral creep measurements at fixed reference points so small growth trends are visible; and a documented creep-life consumption tracking method tied to measured TMT.
Do not rely on the fact that the coil 'ran 11 years last time' - that interval was governed by an uncontrolled temperature excursion, not by design margin, and the next excursion may be faster.