01 Why Two Similar Grades Exist: The Metallurgical Story
In the early 1990s, the global power generation industry confronted a fundamental problem: the next generation of coal-fired power plants — designed to operate at steam temperatures above 600 °C to improve thermal efficiency and reduce CO&sub2; output per megawatt — was out-running the creep rupture capability of the then-dominant 9Cr-1Mo-V steel known as P91 and F91.
F91 had already been a landmark achievement when introduced in the 1970s as an improved successor to F9 (plain 9Cr-1Mo). The addition of vanadium, niobium, and nitrogen to the 9%-chromium base dramatically improved long-term creep rupture strength through precipitation hardening of MX-type carbides and nitrides. But at metal temperatures above 593–600 °C, the MX precipitates coarsen, the lath-martensite microstructure recovers, and creep life drops steeply.
The solution — developed collaboratively by Japanese steelmakers (NKK and Sumitomo), the German VdEW programme, and the Electric Power Research Institute (EPRI) in the United States — was to replace approximately half of the molybdenum in F91 with tungsten, add a small boron micro-addition for grain boundary stability, and fine-tune the vanadium and niobium levels. The resulting composition was designated NF616 in Japan and standardised internationally as Grade 92 / P92 / F92 under ASTM A335, ASTM A182, and ASME SA182 respectively. The European equivalent is EN X10CrWMoVNb9-2 (1.4901).
02 Chemical Composition: Where F92 and F91 Diverge
Both grades share a 9% chromium martensitic backbone but diverge significantly in transition metals and micro-additions. The table below summarises the full composition ranges per ASTM A182/A182M.
| Element | F91 Range | F92 Range | Significance of Difference |
|---|---|---|---|
| Carbon (C) | 0.08–0.12% | 0.07–0.13% | Similar; controls martensite hardness and M₃₃C₆ carbide density |
| Chromium (Cr) | 8.0–9.5% | 8.5–9.5% | Oxidation / steam resistance; identical function in both grades |
| Molybdenum (Mo) | 0.85–1.05% | 0.30–0.60% | Reduced ~60% in F92; partially replaced by tungsten for superior high-temperature stability |
| Tungsten (W) | — (absent) | 1.50–2.00% | Primary differentiator. W diffusivity at 600°C is 4–5× lower than Mo, dramatically slowing precipitate coarsening and lath recovery |
| Vanadium (V) | 0.18–0.25% | 0.15–0.25% | MX carbonitride former; creep strengthening through fine precipitates |
| Niobium (Nb) | 0.06–0.10% | 0.04–0.09% | MX carbonitride former; austenite grain refinement |
| Nitrogen (N) | 0.030–0.070% | 0.030–0.070% | Stabilises MX nitrides; same role in both grades |
| Boron (B) | — (absent) | 0.001–0.006% | Critical F92 micro-addition. Segregates to prior-austenite grain boundaries; retards M₃₃C₆ coarsening; extends creep life at 600°C+ |
| Silicon (Si) | ≤ 0.50% | ≤ 0.50% | Deoxidiser; low to avoid δ-ferrite formation |
| Manganese (Mn) | 0.30–0.60% | 0.30–0.60% | Identical in both grades |
| Phosphorus (P) | ≤ 0.020% | ≤ 0.020% | Temper embrittlement control |
| Sulfur (S) | ≤ 0.010% | ≤ 0.010% | Same |
The boron content window in F92 (0.001–0.006 wt%) is narrow. Excess boron causes grain coarsening during normalising and degrades impact toughness below the minimum Charpy requirement. Always request heat analysis alongside product analysis when procuring F92 forgings, and verify that the boron value is within the ASTM A182 range — not simply "present" or "trace."
03 Mechanical Properties: Room Temperature and Elevated
At ambient temperature, F92 and F91 post comparable strength levels — intentional, since alloy designers wanted F92 to slot into existing code frameworks without requiring re-certification of component wall thicknesses. The real divergence occurs at elevated temperature.
| Property | F91 Min / Typical | F92 Min / Typical | Notes |
|---|---|---|---|
| Tensile Strength (UTS) | 585 MPa min | 620 MPa min | F92 marginally higher; both achieve 650–750 MPa typical |
| 0.2% Proof Stress (Rp0.2) | 415 MPa min | 440 MPa min | Near-identical yield floor in practice |
| Elongation (A%) | ≥ 20% | ≥ 20% | Same minimum; both achieve 22–26% typical |
| Reduction of Area (Z%) | ≥ 45% | ≥ 45% | Same |
| Charpy Impact Energy (RT) | ≥ 41 J | ≥ 41 J | F92 can show lower absorbed energy if boron is at upper limit |
| Hardness | ≤ 248 HBW | ≤ 250 HBW | Near-identical after PWHT; maintains weldability compatibility |
| UTS @ 600°C | ~350–380 MPa | ~380–420 MPa ↑ Better | F92 advantage starts to emerge |
| Rp0.2 @ 600°C | ~230–250 MPa | ~270–290 MPa ↑ Better | Significant difference at this temperature |
04 Creep Rupture Strength: The Critical Differentiator Above 600°C
Tensile strength is tested over minutes. Creep rupture strength is measured over tens of thousands of hours, extrapolated to 100,000 hours (approximately 11.4 years of continuous service). For power plant pressure components, this long-term parameter is what design codes — including ASME BPVC Section II Part D, EN 13480, and TRD — use to set allowable stress values at elevated temperature.
Why tungsten slows creep: the physical mechanism
Creep in tempered martensitic steels is controlled by dislocation climb along sub-grain boundaries, where the rate-limiting step is solid-state diffusion of strengthening atoms. At 600 °C, tungsten atoms have a diffusivity approximately 4–5 times lower than molybdenum atoms. This means W-bearing precipitates (Laves phase Fe₂W) grow more slowly, maintain a finer distribution for longer, and continue obstructing dislocation motion for a greater fraction of the component's service life.
Simultaneously, the boron micro-addition in F92 dramatically reduces the rate of M₃₃C₆ carbide coarsening at prior-austenite grain boundaries — a process that otherwise creates local soft zones that act as preferred creep crack initiation sites in long-service F91 components.
Field experience with F91 components — particularly thick-section forgings such as valve bodies, main steam pipe bends, and turbine bypass valve chests — has revealed Type IV cracking: premature failure in the fine-grained heat-affected zone (HAZ) of welds at lives significantly shorter than predicted by uniaxial creep data. This failure mode becomes more frequent in components operating above 575 °C persistently.
F92 is not immune to Type IV cracking, but its slower microstructural recovery rate delays onset and widens defect tolerance. For new USC plant designs targeting 600 °C+ steam, specifying F92 forgings for USC service at critical weld locations substantially reduces inspection burdens over a 25–40-year plant life.
05 Weldability, Preheat & Post-Weld Heat Treatment
Both F91 and F92 are weldable with appropriate procedure controls but neither tolerates improvisation. Both grades undergo a martensitic transformation during cooling from the austenite range, making untempered martensite — and hence cold cracking — a real risk without strict preheat and PWHT control.
| Parameter | F91 | F92 |
|---|---|---|
| Min Preheat Temperature | 200°C (thin section) to 250°C | 200°C (thin section) to 250°C |
| Max Interpass Temperature | 300°C | 300°C |
| Pre-PWHT Cooling | Cool to 80–100°C to ensure full martensitic transformation before entering furnace | Same — do not go directly from weld temperature to PWHT. Risk: retained austenite that does not temper, then transforms to hard fresh martensite in service |
| PWHT Temperature Range | 730–780°C per code; typically 745–765°C | 730–780°C per code; typically 750–770°C preferred for F92 |
| PWHT Hold Time | 1 hr per 25mm wall thickness, minimum 1 hr | Same minimum; longer soak recommended for heavy sections (>75mm wall) |
| Filler Metal (GTAW root) | ER90S-B9 matching filler per AWS A5.28 | Matching F92 filler with W addition (B9W-type); verify to current AWS A5.28 classification before ordering |
| Max HAZ Hardness Post-PWHT | ≤ 250 HV10 typical | ≤ 265 HV10 typical |
| Dissimilar Metal Welds | Ni-alloy buttering layer (Alloy 82/182 or ERNiCrFe-2) required at junctions with austenitic stainless steel | Same protocol. Note: F92 CTE ≈ 12×10⁻⁶ K⁻¹ vs 316SS ≈ 18×10⁻⁶ K⁻¹ — thermal fatigue at cycling dissimilar joints is a serious failure risk without proper butter design |
Failing to allow the weld assembly to cool below the martensite finish temperature (Mf ≈ 80–100 °C) before initiating PWHT is the most frequently encountered fabrication error on F91 and F92 projects. This results in retained austenite that transforms to untempered martensite during service, initiating premature cracking. Ensure your WPS explicitly mandates the intermediate cool-to-temperature step, and verify that your heat treatment contractor has confirmed understanding of this requirement in writing before work commences.
06 Application Map: Where Each Grade Belongs
The grade boundary is defined by operating temperature of individual components — not by a plant-wide label. Even in a USC plant, low-temperature system segments correctly use F91, F22, or F11 forgings. The decision matrix below covers the most common component locations.
- Main steam metal temperature exceeds 600 °C (USC plant design)
- Reheat steam temperature exceeds 620 °C (Advanced USC, A-USC)
- Main steam stop valves, control valves, bypass valve chests at USC plants
- Thick-section forgings (>100mm wall) with >30-year life targets at 580–600°C
- HP turbine casing bolting above 580 °C where creep relaxation is unacceptable
- ASME allowable stress for F91 is insufficient at the design temperature
- Seamless rolled rings for large-bore hot reheat stop valves (>DN400) in USC service
- Any weld location with historical Type IV cracking risk in F91
- Main steam temperature is 540–593 °C (supercritical or high-subcritical)
- Continuous metal temperature stays below 580 °C throughout plant life
- Retrofit or life-extension of existing F91 plant — code continuity preferred
- Cold reheat piping, feedwater heater shells, extraction steam components
- Valve bonnets and stems at 500–580 °C intermediate-temperature service
- F92 supply chain is not established in the project region; lead time is critical
- Repair forgings where property matching with original F91 system is mandatory
A plant is supercritical when steam pressure exceeds 22.1 MPa (the thermodynamic critical point of water); ultra-supercritical (USC) when steam temperature exceeds 600 °C; and advanced ultra-supercritical (A-USC) for designs targeting 700 °C+. The 700 °C A-USC regime requires nickel-base superalloy forgings (e.g. Inconel 740H) and lies beyond both F91 and F92. F92's productive territory is the 600–625 °C band — the dominant window for USC power plants built globally across the 2010s and 2020s.
07 Cost and Supply Chain Considerations
F92 carries a genuine cost premium over F91, and understanding its sources helps procurement teams build accurate project budgets and negotiate effectively with forging suppliers.
Raw material cost — the tungsten factor
The 1.5–2.0 wt% tungsten content in F92 is the primary cost driver. Tungsten prices are historically volatile and supply is geographically concentrated. The net W addition — partially offset by the reduced Mo content — typically adds 8–15% to base alloy material cost compared to equivalent-section F91 steel. For forgings above 5 tonnes, this translates to a material cost delta of tens of thousands of dollars per component.
Processing and heat treatment premium
F92 billets require tighter normalising and tempering control due to the narrower acceptable microstructural window. Boron content requires precise furnace atmosphere control (no contamination from previous alloy loads), and the W-Mo balance must be verified by heat analysis before forging commences. These controls add approximately 5–10% to processing cost compared to an equivalent F91 forging.
Lead time and global supply
F91 is among the most widely-stocked alloy forging grades globally. F92 has a narrower supplier base and lower ex-stock availability for large forgings in many regions. At Jiangsu Liangyi, standard lead times for F92 open-die forgings and seamless rolled rings are 20–35 working days from order confirmation — comparable to F91 at our facility due to our dedicated heat-certified steel inventory — but buyers sourcing from the broader market should confirm availability early in the project schedule.
Lifecycle economics — where F92 pays back
For new-build USC plants, the lifecycle cost calculus generally favours F92 at the hot end of the steam cycle despite the higher initial component cost. Longer inspection intervals (reduced Type IV cracking risk), fewer forced outages from component distress, and extended time to first major repair — particularly for main steam valves and thick-section pipe fittings — typically deliver total lifecycle savings that outweigh the procurement premium over a 30-year plant life. For retrofit projects on existing plant operating below 600 °C continuously, this justification weakens considerably and F91 remains the correct economic choice.
08 Grade Selection Checklist: F92 vs F91
Work through each question in order. The first definitive "yes" determines your grade specification for that component location.
| # | Question | If YES — Specify |
|---|---|---|
| 1 | Does the component see sustained metal temperature above 600 °C during normal operation? | F92 |
| 2 | Is the wall section greater than 100mm and service life target greater than 30 years at 580–600 °C? | F92 |
| 3 | Does ASME Section II Part D allowable stress for F91 at the design temperature result in wall thickness exceeding the compact-design envelope? | F92 |
| 4 | Is the component at the HP turbine inlet or in hot reheat piping of a plant with reheat steam >610 °C? | F92 |
| 5 | Is this a repair or replacement forging for a system originally constructed in F91 with sustained metal temperature below 580 °C? | F91 |
| 6 | Is the component in cold reheat, extraction steam, or auxiliary steam at temperatures below 550 °C? | F91 (or F22 / F11 at lower temps) |
| 7 | Is the project on a tight capital budget and metal temperature confirmed below 590 °C for the full design life? | F91 |
| 8 | Is the component in a plant where F92 PWHT capability is not established in the maintenance team? | Consider F91 and plan F92 upskill programme |
09 F92 and F91 Forged Products from Jiangsu Liangyi
As an ISO 9001:2015 certified open-die forging manufacturer with over 25 years of experience and an 80,000 m² production facility in Jiangyin, China, Jiangsu Liangyi manufactures both grades across the full product range — from single prototype pieces to large-volume EPC supply contracts. All products are supplied with EN 10204 Type 3.1 Mill Test Certificates as standard — issued and validated by our own ISO 9001:2015-certified quality department. EN 10204 Type 3.2 certificates, which require co-signature by an independent third-party inspector, are available when customers arrange and fund their own inspection body (SGS, BV, TÜV Rheinland, DNV, Lloyd's Register, ABS, etc.).
ASTM A182 F92 Forging Products
- Valve bodies — main steam stop, control, bypass valves for USC power plants
- Valve bonnets, closures, stems, seat rings — precision-machined to drawing
- Seamless rolled rings — OD up to 6 m; flanges, casing rings, seal rings
- Forged pipes and sleeves — main steam and hot reheat piping systems
- Disc and block forgings — valve disc blanks, hubs, housings up to Ø3 m
- Forged bars — round Ø50–2000 mm; square, flat, hollow bar
- Standards: ASTM A182 / ASME SA182 / ASME SA336 F92
ASTM A182 F91 Forging Products
- Same full product range as F92 — all component types and sizes
- Standards: ASTM A182 / ASME SA182 / ASME SA336 F91
- A335 Grade P91 / SA335 Grade P91 for pipe fittings
- EN X10CrMoVNb9-1 (1.4903) European equivalent supplied
- Manufactured to ASTM, ASME, EN and API technical standards as specified per individual purchase order
- NDE: UT per ASTM A388 / EN 10228-3; MT per EN 10228-1
- Products can be manufactured to NACE MR0175 / ISO 15156 material technical requirements when specified at order stage — note: Jiangsu Liangyi holds ISO 9001:2015 certification only; we do not hold API product licences or ASME stamps
What we hold: Jiangsu Liangyi Co., Limited is certified to ISO 9001:2015 Quality Management System by an accredited certification body. This covers our internal quality management processes.
What we manufacture to (not certifications we hold): We manufacture products in accordance with ASTM A182, ASME SA182, ASME SA336, EN, DIN, API, and NACE technical standards as specified in individual purchase orders. Compliance with these technical standards is verified through in-house and third-party testing per the applicable standard — it does not mean we hold API product licences, ASME stamps, PED Notified Body approvals, or other third-party product certifications. Buyers whose projects require such approvals should confirm requirements with their project team before ordering.
Third-party inspection: We fully cooperate with and welcome inspection by SGS, Bureau Veritas (BV), TÜV Rheinland, DNV, Lloyd's Register, ABS, and others. EN 10204 Type 3.2 certificates are issued when customers arrange and fund the inspection body.
10 Conclusion
The ASTM A182 F92 vs F91 decision reduces to a single thermal threshold: 600 °C. Below it, F91 is proven, widely available, and economically rational for supercritical power plant forgings. Above it, the tungsten micro-alloying and boron grain-boundary stabilisation in F92 provide a creep rupture advantage of approximately 40–50% at 600 °C over 100,000 hours — an advantage that no increase in wall thickness or inspection frequency can replicate in F91 at this temperature.
For engineers specifying forgings for ultra-supercritical power plants at or above 600 °C main steam temperature, F92 is the technically correct specification. For supercritical plant operating below that threshold, F91 continues to serve reliably and remains the correct economic choice. The two grades are complements, not competitors — each suited to a distinct thermal regime.
At Jiangsu Liangyi, we produce both grades across the full range of open-die forging and seamless ring rolling processes, under the same ISO 9001:2015 quality system. Our range of F92 forged components — from valve bodies and seamless rings to forged bars and disc forgings — is detailed on our product page, with engineering support available for material selection, code compliance, and weld procedure review.
Specify F92 or F91 for your project?
Our engineering team reviews your operating parameters and provides grade recommendations with a formal quotation within 24 hours.