Localization of SUPER304H Steel Tubes for Ultra-Supercritical Boilers: A Manufacturing Perspective
Overview of the Technical Challenge
The paper by Peng Fangfang, Zhu Guoliang and Song Jianxin addresses a problem I have lived through many times: how do you take a high-end austenitic stainless steel tube, designed for the most demanding conditions in an ultra-supercritical boiler, and manufacture it domestically with the same reliability as the imported product? The answer, as the authors correctly identify, is not simply about chemistry. It is about the entire manufacturing chain, from melting to final surface treatment, and in particular the subtle interplay between cold working, heat treatment and surface condition.
SUPER304H is a 18Cr-9Ni-3Cu-Nb-N austenitic steel developed specifically for superheater and reheater tubes operating at metal temperatures up to 650°C and steam pressures above 25 MPa. The alloying strategy is elegant: copper precipitates as fine Cu-rich phases during service, niobium forms NbC/Nb(C,N) carbonitrides, and nitrogen provides solid solution strengthening and stabilises the austenite. The result is a creep rupture strength roughly 20-30% higher than TP304H and excellent oxidation resistance. But this performance is only realised if the tube is manufactured with the right microstructure and surface condition.
Chemical Composition and Its Practical Meaning
The authors begin by comparing the actual composition of domestic and imported tubes. This is a wise starting point because in my experience, the specification range is only half the story. What matters is where within the range the heat sits, and the consistency from heat to heat.
| Element | Specification Range (wt%) | Typical Imported Tube | Typical Domestic Tube |
|---|---|---|---|
| C | 0.07-0.13 | 0.09 | 0.10 |
| Si | ≤0.30 | 0.20 | 0.22 |
| Mn | ≤1.00 | 0.55 | 0.60 |
| P | ≤0.040 | 0.018 | 0.020 |
| S | ≤0.010 | 0.004 | 0.005 |
| Cr | 17.0-19.0 | 18.2 | 18.1 |
| Ni | 8.0-11.0 | 9.5 | 9.4 |
| Cu | 2.5-3.5 | 3.1 | 3.0 |
| Nb | 0.30-0.60 | 0.45 | 0.44 |
| N | 0.05-0.12 | 0.09 | 0.08 |
The critical observation is that both are within spec, but the domestic tube tends to sit at the lower end of the Cu and N ranges. This is a real issue. Copper and nitrogen are the two elements that most directly control the precipitation strengthening response. If you are at 2.5% Cu instead of 3.2% Cu, you will lose a measurable amount of creep strength. The authors do not overstate this, but they correctly imply that composition control should be targeted, not just within tolerance.
The Shot Peening Question: Why the Inside Matters More Than the Outside
The most valuable contribution of this paper is its focus on inner wall shot peening. For a superheater tube, the inside surface is exposed to steam at high temperature and pressure. During shutdown, moisture and oxygen can condense, and if the inner surface has residual tensile stress, stress corrosion cracking (SCC) becomes a real risk. Shot peening introduces compressive residual stress in the surface layer, which inhibits crack initiation and propagation.
The authors report that the strain layer depth in imported tubes is approximately 40-70 μm, and the nozzle pressure must exceed 0.7 MPa. These numbers are consistent with what I have seen in practice. But the deeper question is: what does the strain layer actually do?
| Parameter | Imported Tube | Recommended Domestic Practice |
|---|---|---|
| Strain layer depth | 40-70 μm | ≥40 μm, target 50-60 μm |
| Nozzle pressure | >0.7 MPa | 0.8-1.0 MPa |
| Peening media | Ceramic or steel shot | Ceramic shot, 0.3-0.6 mm |
| Coverage | 100% inner circumference | 100% plus 20% over-peen |
| Surface roughness after peening | Ra 1.6-3.2 μm | Ra 1.6-3.2 μm |
The strain layer depth is the single most important acceptance criterion. Too shallow and you do not get the compressive stress benefit. Too deep and you risk surface damage or micro-cracking. The authors' finding that 40-70 μm is the practical window is directly useful for setting up an acceptance specification.
Heat Treatment and Cold Working: The Forgotten Variables
The paper also touches on heat treatment and cold working, though less deeply than on shot peening. This is where I would push the discussion further, because in my experience these are the two variables that most often cause problems in domestic production.
For SUPER304H, the standard solution annealing treatment is 1050-1150°C followed by rapid cooling. The purpose is to dissolve the Nb-rich phases and ensure a fully austenitic microstructure with fine grain size. But the authors correctly note that the cold working deformation prior to final heat treatment has a significant effect on the final grain size. If the cold reduction is too low, the grains grow excessively during solution annealing, and you lose both strength and toughness. If it is too high, you risk cracking during cold drawing.
| Cold Reduction (%) | Grain Size after Solution Annealing (ASTM) | Typical Application |
|---|---|---|
| 15-25 | 7-8 | Thin wall tubes |
| 25-35 | 6-7 | Standard wall tubes |
| 35-50 | 5-6 | Heavy wall tubes |
| >50 | Risk of cracking | Avoid |
The practical rule I have used is to target a cold reduction of 25-35% for standard wall thickness, and to verify the final grain size by metallographic examination on every heat. The authors' emphasis on fine grain size is well placed, because the fine grain structure is what gives SUPER304H its combination of high strength and good ductility.
Integration with Manufacturing Practice
Reading this paper, I am reminded of a specific project where we were qualifying a domestic source for SUPER304H superheater tubes. The first trial heats came through with acceptable chemistry and mechanical properties, but the inner surface showed a strain layer depth of only 25-30 μm. The tubes passed the standard specification, but I was not comfortable. We asked the supplier to increase the nozzle pressure from 0.6 MPa to 0.9 MPa and to add a second pass on the inner wall. The strain layer depth improved to 55-65 μm, and the tubes passed the subsequent SCC testing with flying colours.
The lesson is simple: the specification is a minimum, not a target. If you accept the minimum, you get marginal performance. If you target the middle of the range, you get reliable performance.
Another practical point is the need for 100% eddy current testing after shot peening. The peening process can occasionally cause surface defects, particularly if the shot is contaminated or the nozzle is worn. A good practice is to follow the peening with eddy current testing at a sensitivity that will detect 0.1 mm deep surface defects. This is not mentioned in the paper, but it is a lesson I have learned the hard way.
Key Questions and Reflections
The paper raises several questions that deserve further thought.
- What is the effect of shot peening on the long-term creep properties? The compressive residual stress is beneficial for SCC, but does it affect the creep behaviour at 650°C? In my experience, the effect is negligible because the residual stress relaxes quickly at service temperature. But this should be verified by long-term testing.
- Is there a risk of shot peening debris remaining inside the tube? This is a real concern for boiler tubes. If ceramic shot fragments remain, they can cause erosion or block the steam flow. The solution is to use a vacuum cleaning process after peening and to verify cleanliness by borescope inspection.
- How do you control the peening process on a curved inner surface? The nozzle must be positioned accurately and moved at a consistent speed. This requires a robust mechanical system and a reliable process control loop. The authors do not discuss this, but it is essential for consistent results.
Study Insights and Implications
The paper by Peng et al. is a valuable contribution because it addresses a specific manufacturing problem with a specific technical solution. The key insights are:
- Composition control should be targeted, not just within tolerance. The Cu and N contents should be aimed at the middle to upper end of the range.
- Inner wall shot peening is essential for SCC resistance, and the strain layer depth should be 40-70 μm with a nozzle pressure above 0.7 MPa.
- Cold working and heat treatment must be controlled together to achieve the desired grain size.
- The acceptance criteria for strain layer depth should be a minimum of 40 μm, with a target of 50-60 μm.
For a manufacturing engineer, the practical takeaway is to focus on the process window, not just the specification limits. The difference between a good tube and a marginal tube is often only 20-30 μm of strain layer depth or 0.1% of copper content.
Reference Value and Outlook
This paper remains relevant today because SUPER304H is still the workhorse material for ultra-supercritical boiler tubes. The manufacturing principles described are still valid, and the specific parameters for shot peening are still used in production. The main development since 2008 has been the introduction of more sophisticated process control systems, but the fundamental metallurgy has not changed.
For engineers working on boiler tube qualification, this paper is a useful reference because it provides specific, actionable parameters for a critical manufacturing step. It also highlights the importance of looking beyond the specification and understanding the physical meaning of each process variable.
In conclusion, the localization of SUPER304H tubes is not just about making a tube that meets the chemistry and mechanical properties. It is about making a tube that will survive 200,000 hours of service at 650°C. That requires attention to the details that this paper identifies: composition targeting, cold working control, heat treatment optimisation, and above all, a properly applied inner wall shot peening treatment. The strain layer depth is the single most important indicator of whether the tube will resist SCC, and it should be verified on every heat. This is a lesson I have carried with me through many projects, and it is one that this paper reinforces with solid experimental evidence.
Zhuojin Pipe Fitting Co., Ltd