Zhuojin Pipe Fitting Co., LtdQuality Identification and Performance Evaluation of P91 Seamless Steel Pipes
Literature Overview
This paper, authored by He Biao, Zhao Qingquan, Xiao Gongye, Wang Guoliang, Zhang Junping, and Qin Libo from the Technical Center of Tianjin Pipe Group Co., Ltd., was published in Sichuan Metallurgy (Vol. 38, No. 6, 2016, pp. 22–26). The study addresses a critical and recurring problem in the power generation and petrochemical industries: the wide variation in quality among P91 seamless steel pipes on the market, which can lead to serious failures during service. The authors conducted a systematic investigation into heat treatment regimes, hardness, microstructure, chemical composition, impact energy, inclusion content, and processability to establish practical methods for discriminating superior from inferior P91 pipe material.
Core Technical Content
P91 is a 9Cr-1Mo-V-Nb-N steel that has become the workhorse material for ultra-supercritical boiler tubes, reheater tubes, and main steam lines in modern coal-fired power plants operating at temperatures up to 620°C. Its excellent high-temperature strength and oxidation resistance stem from a combination of solid solution strengthening (Cr, Mo), precipitation strengthening (NbC, VC, MX-type carbides), and fine-grain strengthening. However, the very same alloying elements that give P91 its superior properties also make it susceptible to manufacturing inconsistencies, especially during the heat treatment stage.
The paper discusses the critical importance of the normalizing and tempering cycles applied to P91 pipes. Typical heat treatment parameters for P91 seamless pipes are summarized below:
| Parameter | Normalizing | Tempering |
|---|---|---|
| Temperature | 1020–1080°C | 750–770°C |
| Soak time | 20–40 min per 25 mm thickness | 2.0–3.0 h |
| Cooling rate | Furnace cool to 750°C, then air cool | Furnace cool to 500°C, then air cool |
| Target hardness | 180–250 HV | 180–250 HV |
| Target microstructure | Fine ferrite + tempered martensite | Fine ferrite + tempered martensite |
The authors emphasize that deviations in these parameters can produce microstructures that appear superficially acceptable but possess hidden weaknesses. For example, an insufficient normalizing temperature may leave coarse grain boundaries and untempered martensite islands, while excessive tempering temperature can lead to excessive carbide coarsening and a drop in creep strength.
Quality Discrimination Methods
The study proposes a multi-faceted approach to quality identification, which I find particularly valuable for incoming inspection and supplier qualification programs:
Chemical Composition Verification
P91 has tight compositional requirements. Key elements and their critical ranges are:
| Element | Minimum (%) | Maximum (%) | Notes |
|---|---|---|---|
| C | 0.08 | 0.12 | Higher C leads to brittleness |
| Cr | 8.50 | 9.50 | Below 8.5% loses oxidation resistance |
| Mo | 0.85 | 1.05 | Below 0.85% reduces creep strength |
| V | 0.18 | 0.28 | Controls MX carbide density |
| Nb | 0.06 | 0.10 | MX carbide stabilizer |
| N | 0.030 | 0.070 | Excess N causes nitride precipitation |
The paper notes that some suppliers may pass chemical analysis while still producing inferior material due to poor homogeneity or segregation. Spectrographic analysis at multiple positions along the pipe length is recommended rather than relying on a single sample.
Hardness and Microstructure Analysis
Hardness testing is the most straightforward screening tool. The target range of 180–250 HV is critical. Values below 180 HV suggest over-tempering or insufficient alloying, while values above 250 HV indicate possible untempered martensite or inadequate normalizing. The authors recommend testing at both the inner and outer surfaces of the pipe wall, as differential cooling during heat treatment can create hardness gradients.
Microstructural examination reveals more subtle issues. A proper P91 microstructure should show fine, evenly distributed tempered martensite with fine MX-type carbides (Nb(C,N) and VC) and a moderate density of MC carbides (M23C6, M6C). Coarse grain boundaries, excessive carbide stringers along prior austenite grain boundaries, or the presence of retained austenite are all indicators of inadequate processing.
Impact Energy and Inclusion Assessment
Charpy V-notch impact testing at 20°C is a sensitive indicator of microstructural quality. The paper highlights that inferior P91 material may show impact energies below 200 J even when hardness is within specification. Inclusion analysis, particularly for oxide inclusions, is also critical. Large or chain-like oxide inclusions can act as crack initiation sites during service, especially under cyclic thermal loading.
Engineering Practice Implications
In my experience, the failure of P91 components in service is frequently traced back to material quality issues rather than design or fabrication errors. A practical quality assurance protocol based on this literature would include:
- Incoming chemical verification at three positions along the pipe length (head, middle, tail) using optical emission spectrometry.
- Hardness mapping at 5 locations along the pipe circumference at both ID and OD surfaces, with results plotted as a contour map.
- Metallographic examination of a transverse section at the head end to assess grain size (target ASTM 8–10), carbide distribution, and absence of retained austenite.
- Charpy impact testing at both 20°C and –20°C to verify ductility and toughness margins.
- Inclusion rating per ASTM E45, with a maximum allowance of Type 1, 2, 3, 4 inclusions at level 1.5.
A notable engineering case that illustrates the importance of these checks involved a reheater tube rupture in a 1000 MW ultra-supercritical unit. Post-failure analysis revealed that the tube material, while passing basic chemical and hardness checks, contained a high density of chain-like oxide inclusions that initiated cracks during thermal cycling. This case underscores the value of the multi-parameter discrimination approach advocated in this paper.
Key Reflections and Study Insights
The most valuable contribution of this paper is its emphasis on systematic, multi-parameter quality evaluation rather than reliance on any single test. In practice, many procurement specifications focus heavily on chemical composition and mechanical properties while neglecting microstructural and inclusion assessments. This creates a false sense of security. I believe that incorporating the quality identification framework from this paper into supplier qualification protocols would significantly reduce the risk of field failures.
Furthermore, the paper's discussion of heat treatment sensitivity highlights the need for rigorous process control during manufacturing. Even minor deviations in furnace temperature uniformity, cooling rate, or soak time can produce microstructural differences that are difficult to detect by conventional testing but have significant consequences for long-term service performance. This reinforces the importance of process qualification and periodic process audits at pipe manufacturers.