Welding Construction Optimization for P91 Heat-Resistant Steel Pipeline
Why P91 Is Both Valuable and Difficult
P91 is a 9Cr-1Mo martensitic heat-resistant steel developed for high-temperature, high-pressure service in power plants and petrochemical facilities. Its creep strength at elevated temperatures is excellent, which is why it has largely replaced P22 in many applications. But P91 is also one of the most difficult steels to weld in the entire pressure piping family. The metallurgy is unforgiving: the steel transforms to martensite on cooling from the welding temperature, and the resulting microstructure is hard, susceptible to hydrogen-induced cold cracking, and prone to HAZ softening and Type IV cracking in long-term service. The paper reports on a real project — an integrated ethylene module construction project — where the welding of P91 pipelines was optimised through a systematic five-dimensional strategy covering personnel, equipment, materials, methods and environment. The results are impressive: the non-destructive testing pass rate improved from 96% to 99.5%.
The Five-Dimensional Optimization Strategy
The paper adopts a "人、机、料、法、环" (man, machine, material, method, environment) framework, which is the Chinese equivalent of the 5M approach in quality management. Let me break down what each dimension means in practice for P91 welding.
| Dimension | Key Actions | Why It Matters |
|---|---|---|
| Man (人) | Strict welder qualification review, verify welder certification against the specific process | P91 welding requires high skill; an unqualified welder can ruin a joint in seconds |
| Machine (机) | Verify welding equipment calibration, ensure proper arc characteristics and shielding gas flow | Inconsistent arc energy leads to inconsistent heat input and variable microstructure |
| Material (料) | Verify filler metal grade, check hydrogen content, control storage and handling | Low-hydrogen consumables are essential; moisture pick-up is a killer |
| Method (法) | Optimise preheat, introduce temper bead technique, standardise PWHT procedures | Controls cooling rate, reduces hardness, tempers the HAZ |
| Environment (环) | Control ambient temperature, humidity and wind during welding | Prevents moisture contamination and rapid cooling |
Welder Qualification
The first step is rigorous welder qualification. P91 welding is not something you hand to a general pipe welder. The welder must be qualified specifically for P91, with the specific process, the specific filler metal and the specific position. I have seen projects where a welder qualified on carbon steel was allowed to weld P91 because the company thought "steel is steel." That is a recipe for disaster. The qualification test must include the required hardness testing and metallographic examination of the HAZ, not just a bend test and a radiograph.
Preheat and Interpass Control
Preheat is critical for P91. The steel transforms to martensite at a relatively high temperature, and if the cooling rate is too fast, the martensite is hard and susceptible to cold cracking. The paper mentions optimising the preheat process, which typically means a preheat temperature in the range of 150–250°C for P91, with interpass temperature controlled to a maximum of 300°C or so. The preheat must be applied uniformly, and the temperature should be checked with contact pyrometers or infrared thermometers at the specified locations. I have seen too many cases where preheat was applied to the pipe but not to the fitting, or where the preheat was allowed to decay before the welding started. The result is a cold joint that cracks either immediately or within a few days.
Temper Bead Technique
The introduction of the temper bead technique is a smart move. The idea is that the final weld pass is deposited in a way that tempers the HAZ of the previous passes, reducing the hardness and improving the toughness. This is particularly useful where PWHT is difficult or where the joint geometry makes full PWHT impractical. The temper bead technique requires careful control of bead placement, heat input and interpass temperature, and it should only be performed by welders who have been specifically trained in the technique. The paper reports that this technique, combined with standardised PWHT, contributed to the significant improvement in NDT pass rate.
Post-Weld Heat Treatment
PWHT for P91 is a delicate operation. The standard requirement is typically 730–780°C for a specified holding time, but the actual temperature and time must be carefully controlled to achieve the right balance of hardness reduction and creep strength retention. Too low a temperature and the hardness remains too high; too high a temperature and the creep strength is compromised. The paper mentions standardising the PWHT procedures, which means ensuring that the heating rate, holding temperature, holding time and cooling rate are all controlled and recorded. The temperature should be measured with thermocouples attached to the weld, not just the pipe surface, and the number and placement of thermocouples should be specified in the procedure.
What the NDT Results Mean
The improvement from 96% to 99.5% NDT pass rate is significant, but let me put it in perspective. A 96% pass rate means that 4 out of every 100 welds had a defect that required repair. On a large module project with hundreds of P91 welds, that is a lot of repairs. Each repair involves grinding out the defect, re-welding, re-inspecting and re-heat-treating, which can take days. The improvement to 99.5% means only 0.5% of welds required repair, which is a dramatic reduction in rework cost and schedule impact. The paper attributes this improvement to the five-dimensional optimisation strategy, and I have no reason to doubt it. The combination of qualified welders, calibrated equipment, low-hydrogen consumables, optimised preheat, temper bead technique and standardised PWHT is exactly the recipe I would prescribe.
Closing Reflections
P91 is a steel that rewards discipline and punishes shortcuts. The paper's five-dimensional strategy is a comprehensive framework that addresses every aspect of the welding process, and the results speak for themselves. My own experience with P91 has taught me a few lessons that are worth repeating. First, never let a welder who is not specifically qualified for P91 touch the material. The qualification must be specific to the process, the position and the filler metal. Second, control the hydrogen. P91 is extremely sensitive to hydrogen-induced cold cracking, and every source of hydrogen — moisture in the flux, moisture in the shielding gas, oil or grease on the bevel, even the humidity in the air — must be controlled. Third, do not rush the PWHT. The heating rate, holding temperature and cooling rate must be controlled and recorded, and the temperature must be verified with thermocouples at the weld. Fourth, inspect with the right methods. RT and UT are both required for P91 welds, and the acceptance criteria should be strict. Finally, remember that P91 welds can crack days or even weeks after welding, so the final inspection should be done after a suitable delay, not immediately after PWHT. The paper's findings are consistent with everything I have learned in thirty years of welding and supplying pipe, and I would recommend the five-dimensional strategy to anyone facing a P91 welding project. It is not the cheapest way to do the job, but it is the only way I know to do the job right.
Zhuojin Pipe Fitting Co., Ltd