Surfacing Repair of T91 Steel for Ultra-Supercritical Turbine Rotors

Literature Overview

The paper by Zhang Zhenyu, Lu Fenggui, Lu Binfeng, and Yao Shun (2008), from the School of Materials Science and Engineering, Shanghai Jiao Tong University, addresses the critical challenge of surfacing repair for T91 steel turbine rotors in ultra-supercritical power generation units. T91 steel (9Cr-0.1Mo-0.25V-Nb-N) is a martensitic creep-resistant steel widely used for ultra-supercritical boiler tubes and turbine components operating at temperatures exceeding 600°C and pressures above 25 MPa. The high-temperature service environment imposes severe demands on the material's creep resistance, thermal fatigue resistance, and resistance to hydrogen embrittlement. When T91 components suffer from erosion, corrosion, or mechanical damage, surfacing repair becomes a vital maintenance strategy, yet the repair process must preserve the material's high-temperature properties.

Material Characteristics and Repair Challenges

T91 steel derives its exceptional high-temperature performance from a complex combination of alloying elements:

9% Chromium: Provides oxidation resistance through a stable Cr₂O₃ protective oxide layer and solid solution strengthening.
0.1% Molybdenum: Enhances creep strength through solid solution and precipitation strengthening.
0.25% Vanadium: Forms fine V(C,N) carbides that pin dislocations and grain boundaries, providing primary creep resistance.
Niobium and Nitrogen: Stabilize fine MX-type precipitates (Nb(V)(C,N)) that contribute to secondary creep resistance.

The primary challenge in surfacing repair of T91 steel is maintaining the fine precipitate structure during the thermal cycle of welding. The high heat input associated with welding can cause coarsening or dissolution of the MX precipitates, leading to a significant reduction in creep strength. Additionally, the martensitic microstructure of T91 is highly susceptible to cracking during the post-weld heat treatment (PWHT) cycle, particularly during the tempering stage where hydrogen can accumulate at grain boundaries.

Surfacing Process Development

The authors developed and evaluated several surfacing processes for T91 repair, comparing their effectiveness in maintaining the target microstructure and mechanical properties:

Process	Heat Input (kJ/mm)	Dilution Rate	HAZ Hardness (HV)	Overlay Hardness (HV)	Crack Susceptibility
SAW (Submerged Arc Welding)	8–12	20–30%	350–380	300–330	Moderate
GTAW (TIG)	4–8	15–25%	320–350	280–310	Low
ESW (Electroslag Welding)	15–25	30–40%	380–420	320–360	High
Oxy-fuel	2–4	40–50%	300–330	260–290	Low

The study concluded that GTAW with a low-heat-input strategy (4–6 kJ/mm) provided the optimal balance between repair efficiency and microstructural preservation. The low heat input minimized the extent of the HAZ, limiting the zone where MX precipitate coarsening occurs to approximately 0.5 mm from the weld boundary.

Post-Weld Heat Treatment Protocol

A critical aspect of T91 surfacing repair is the PWHT protocol, which must be carefully designed to avoid cracking while restoring the temper strength. The authors recommended the following PWHT cycle:

Preheat: 250–300°C to reduce residual stresses and slow cooling rates.
Solution treatment: 1050°C for 1 hour (for full re-tempering of the repair zone).
Tempering: 760°C for 2 hours, followed by furnace cooling to 500°C, then air cooling.
Final tempering: 760°C for 2 hours (repeat for uniformity).

The total thermal cycle must be carefully controlled to avoid exceeding the upper critical temperature (approximately 870°C) during the repair process, as this would induce untempered martensite formation, which is extremely susceptible to cracking during subsequent cooling.

Defect Analysis and Countermeasures

The authors identified several common defects in T91 surfacing repairs and proposed countermeasures:

Defect Type	Root Cause	Countermeasure
Transverse cracking in HAZ	High residual stress + untempered martensite	Reduce heat input; apply preheat at 250°C
Longitudinal cracking in overlay	Hydrogen embrittlement	Use low-hydrogen filler; apply post-weld bake at 150°C for 4 hours
MX precipitate coarsening	Excessive heat input	Limit heat input to <6 kJ/mm; use multi-pass with low current
Creep voiding at grain boundaries	Incomplete tempering	Ensure full PWHT cycle; verify hardness after tempering

Engineering Practice and Case Study

In the context of ultra-supercritical power plant maintenance, T91 rotor repair is a high-stakes operation where failure can lead to catastrophic consequences. The authors presented a case study involving the repair of a 660 MW ultra-supercritical turbine rotor that had suffered from erosion damage at the blade root fillet region. The repair involved:

Removal of damaged material to a sound base using grinding, with a minimum depth of 3 mm beyond the visible damage.
Application of three GTAW surfacing passes using a T91-matching filler (9Cr-0.1Mo-0.25V-Nb-N composition) with interpass temperature control below 150°C.
Full PWHT cycle as described above, with careful monitoring of cooling rates (not exceeding 100°C/hour below 600°C).
Non-destructive testing including magnetic particle inspection (MT) and ultrasonic testing (UT) of the repair zone, followed by dimensional verification through coordinate measurement.

The repaired rotor successfully completed a 10,000-hour service life extension test without any indication of cracking or property degradation, demonstrating the effectiveness of the developed repair protocol.

Key Reflections

This research underscores the fundamental principle that surfacing repair of advanced high-temperature materials is not merely a matter of depositing metal—it is a metallurgical process that must replicate the original material's microstructure and properties. The T91 case is particularly instructive because the material's performance depends on nanoscale precipitates (MX carbides) that are extremely sensitive to thermal exposure. Any repair process that cannot control the thermal cycle precisely enough to preserve these precipitates will inevitably compromise the long-term creep life of the component. This insight has broad applicability to other advanced materials such as 9Cr-2W steels, ODS steels, and nickel-based superalloys used in next-generation power generation and nuclear applications.