New NDE tools facilitate in-service evaluation of P91 condition

The materials traditionally used in powerplant piping systems, such as P22, served well at the nominal 2400-2800-psig/1000F-1050F steam conditions that dominated in coal-fired utility generating facilities from about the late 1950s through the mid-1990s. But the trend to more efficient plants—both conventional steam and combined cycles—required materials capable of operating at higher pressures and temperatures and of accommodating the faster startups and load changes demanded by grid operators in the must-take renewables era.

The 9-Cr martensitic steels P91 and P92 are two materials developed for such demanding service conditions. The former, preferred by US power producers, has been in service here for more than two decades. However, it wasn’t until the gas-turbine bubble (1998-2002) and the proliferation of F-class combined cycles that use of P91 became widespread.

Application of P91 has been challenging for a variety of well-documented reasons—including poor quality control in manufacture and inadequate procedures and quality control in field welding. There have been many incidents of cracking in, and failures of, components made from 9-Cr steels. Thick-section components generally have failed by Type IV cracking associated with welds. Such cracking refers to the premature failure of a welded joint because of an increased rate of creep-void formation in the heat-affected zone (HAZ) attributed to subcritical (intercritical) annealing.  

A possible emerging issue is the sharp drop in creep ductility for both P91 and P92 steels with operating time, reports Dr Ahmed Shibli, director, European Technology Development (ETD), a UK consultancy. The good news: These R&D findings have not yet been confirmed by plant experience. This might reflect the limited use of P91 until recently. The long-term concern is the possibility of base-metal cracking as service life increases.

A problem with Type IV failure, continued Shibli, is that cracking usually starts below the surface and cannot always be detected by traditional nondestructive examination (NDE) procedures—such as replication. In P91, such cracking emerges at the surface near the last leg of its journey, raising safety concerns.

Shibli, recognized globally for his powerplant materials expertise, then added the following concern: Unlike low-alloy ferritic steels operating in the creep regime, with P91 and P92 you cannot easily identify the telltale signs of emerging problems—such as spherodisation and break down of microstructure—even with an optical microscope.

To make matters worse, and life assessment more difficult, he told the editors, it is now fairly well established that high-Cr martensitic steels do not go through the stages of creep-cavity initiation, cavity growth, micro-cracking, and macro-cracking before failure, as do low-alloy steels. They proceed from creep-cavity initiation through cavity growth to failure without warning. This means that for P91 and P92 it’s important to measure and relate cavity density and size, etc, to remaining life.

Yet another difference between traditional low-alloy steels and 9-Cr martensitic steels is that cavity size for the latter can remain in the nanometer range for up to about 70% of a component’s lifetime, thereby making cavity detection and quantification by traditional NDE methods difficult until it may be too late for taking effective remedial steps. The increasing number of P91 failures worldwide has dramatically raised the level of interest in integrity/life assessment, inspection, and monitoring of 9-Cr steels. It is important that the new techniques be sensitive enough to locate nano-size cavities and to quantify their densities/areas/volumes with service time. In the case of Type IV damage, Shibli said, the techniques must be sufficiently sensitive to inspect and monitor sub-surface cavities.

ETD recently completed Phase 1 research and preliminary development of new techniques that qualify for the inspection and monitoring of P91. The consultancy, together with European and Japanese partners, launched this month a two-year Phase 2 initiative: “P91-P92 Inspection and Lifing.” It includes a creep and creep-fatigue testing program on P91 and P92 pipe to confirm and demonstrate the validity of the techniques investigated in Phase 1 and to validate the use of inspection data collected by them for integrity and life assessments.

Here’s a rundown on four of the candidate tools being investigated for inspection and monitoring of 9-Cr martensitic steels:

* Scanning force microscope (SFM) is a portable version of the atomic force microscope long used in medical research but not well known in the power and process industries. After many years of R&D, ETD, through its European collaborators, has developed a portable version of the SFM suitable for inspecting industrial components that may exhibit creep, fatigue, or corrosion damage. It has been used successfully to inspect steam valves, P22 and P91 pipe (Fig 1), steam-turbine rotors (Fig 2), pipe fittings, etc. Fig 3 illustrates the value of SFM in assessing creep cavitation.

1, 2. Scanning force microscope can reveal creep, fatigue, and/or corrosion damage in pipe (left) and rotor disc (right)

1, 2. Scanning force microscope can reveal creep, fatigue, and/or corrosion damage in pipe (left) and rotor disc (right)

3. SFM image at left reveals creep cavitation in creep-damaged P91 piping; at right, cavity depth is determined by a line-scan of the SFM image

3. SFM image at left reveals creep cavitation in creep-damaged P91 piping; at right, cavity depth is determined by a line-scan of the SFM image

The SFM is being used both for nano-level damage detection and microstructure identification. A deep creep cavity is shown in Fig 4. Also, the use of SFM for early-stage (about 20% of the design lifetime) damage detection in P91 has been achieved.

4. SFM image of creep damage in 3-D reveals a deep cavity

4. SFM image of creep damage in 3-D reveals a deep cavity

One of ETD’s long-term goals is to replace all replications with SFM inspections. Data gathering and analysis are fast, unlike that for replication. Cavity-damage data and images are delivered in a flexible format to a laptop connected to the SFM. This enables plotting of information in various configurations—such as cavity size or volume against remaining component life, etc. These images and data then can be transmitted virtually instantly (no image or other processing necessary) to those requiring the information.

* Electrical discharge sampling equipment (EDSE) aims to eliminate the need (1) to remove “boat samples” from thick-section components for full laboratory analysis and (2) to machine out miniature creep, fatigue, or small punch specimens for evaluations of mechanical properties (Fig 5).

5. Electrical discharge sampling equipment removes metal from thick-section components for evaluation of physical and mechanical properties

5. Electrical discharge sampling equipment removes metal from thick-section components for evaluation of physical and mechanical properties

With heat treatment of the base metal and welded components critical for obtaining optimum properties of 9-Cr martensitic steels, EDSE has proved invaluable for assessing the integrity of both newly manufactured and ex-service components were such integrity may be suspect. Use of the method extends beyond P91 and P92. Recently it was used to investigate graphitization in thick-section carbon-steel pipe by cutting out small samples for laboratory analysis. There usually is no need to repair the material, provided wall thickness remains within design limits.

* High-sensitivity ultrasonic probes, under development to detect nano-size sub-surface damage early in life, will be demonstrated. 

* High-accuracy laser-guided portable hardness tester under development will be used to measure the hardness of the parent material, weld, and HAZ (both coarse- and fine-grain regions) on test pipes, and the change in hardness with test duration. The information then will be related to component life.

To learn more about this multi-client industry sponsored project and details regarding participation, contact Shibli directly.

 

Posted in HRSGs |

Comments are closed.

Categories