Charles Henley, a senior mechanical engineer at Black & Veatch, was one of the presenters at the 2017 VMA Technical Seminar held in early March in Nashville. His presentation on supercritical power plants was especially well received because he shed light on some of the critical issues. This report is based on what he said in his session.
THE HISTORICAL CONTEXT
Until WWII, power plants were relatively low heat, and not particularly efficient when compared to today. After WWII in the 1950s and 1960s, however, efficiencies improved when reheating cycles were instituted. Then, in 1957, the first supercritical power plant was built in the U.S: A 125-megawatt plant operating at 4,500 psi with temperatures at 1150°F/1050°F/1000°F (621°C/566°C/538°C)—a double reheating system. Today, more than 200 supercritical units have been built around the world with Japan and South Korea claiming the largest capacities.
Both subcritical and supercritical are Rankine cycle processes, which means they are comprised of a boiler, a turbine and a condenser. All the systems within the plants support those three pieces of equipment, including the power block and the balance of plant and air quality control systems. The main differences between subcritical and supercritical plants are the higher temperature and pressure levels.
There are four major levels of plants (Figure 1). While there are many levels of advanced plants, essentially, they can take on higher temperatures and pressures. At these higher levels, the plants are more efficient and have fewer emissions.
While subcritical plants generate power by boiling water into steam, supercritical systems turn water to steam without boiling. A supercritical steam generator operates at pressures above the critical pressure (where liquid water immediately becomes steam). The term, “supercritical” is given because the pressure is above the “critical point” of 3,200 psig.
MODERN PLANT MAKEUP
Supercritical power plants have four unique systems: boiler feedwater, main steam, hot reheat and startup.
Engineers in these plants select piping material first. Then, they decide what components will be needed and what families of materials can be used for valves and attendant components that will ensure compatibility with the piping system materials.
The extreme high temperatures and pressures in supercritical power units have a huge impact on the choices. The extreme high pressures necessitate thicker materials with a higher stress range, high cyclic fatigue resistance and increased creep resistance. The extreme temperatures also present the potential for fire-side corrosion and steam-side oxidation in the boiler, which is taken into consideration.
Erosion and Corrosion in Feedwater Systems
Erosion is the damage to materials caused by physical processes such as high-speed impinging flows or solid impacts on the surface. In plants, this can be from cavitation, flashing, solid particle erosion or droplet impingement.
Corrosion is chemical or electrochemical attacks on materials, which can create a widespread attack on an operating system resulting in more general or flow-accelerated corrosion. This also can result in localized attacks such as galvanic corrosion, which occurs when two dissimilar metals are in contact or in crevice corrosion or cracking and pitting.
Flow accelerated corrosion (FAC) is a process that degrades carbon steel material. It first showed up in the nuclear industry, and it appears under relatively unique conditions, depending on the chemistry and PH of the water. FAC occurs when normally protective iron-oxides dissolve into the flowing stream. It also occurs around elbows and Ts, which cause turbulence in the flow, as well as in localized high-velocity areas. FAC is a global attack on piping that creates widespread thinning and can cause substantial wall loss over time. In other words, it’s not like a local attack such as pitting or cracking; FAC-caused failures are often sudden and catastrophic.
In the past, failures due to FAC have caused major incidents with loss of life (Figure 2). FAC has not caused as many failures in recent years because a good inspection system is now in place.
FAC damage appears a few different ways: Under single-phase (i.e., water only) conditions, the damaged surface displays a “scalloped” or “orange-peel” surface. This type of surface is conclusive evidence FAC occurred, though depending on conditions, magnification may be required to see the scalloping (Figure 3).
Under two-phase conditions, the surface may show a pattern of dark and light areas known as “tiger stripping,” which is also conclusive evidence of FAC. Wet steam of some type will cause this (Figure 4).
Where It Hits
Piping is susceptible to FAC if 1) the material is carbon steel, 2) there is water or wet steam flowing in the pipes, 3) temperature conditions are within band 200°F to 500°F (93°C to 260°C) and 4) the water is deoxygenated (i.e., service water systems do not experience FAC)
Systems where FAC is of concern are:
- Generally, all the secondary side of a pressurized water reactor and the equivalent boiling water reactor (BWR) systems
- Some BWR auxiliary systems (e.g., residual heat removal may also be susceptible.)
- In fossil plants, condensate and boiler feed systems as well as some extraction steam lines.
- Some auxiliary systems such as building steam
An important point to note is that, while it is basically impossible to eliminate this kind of damage, it can be managed through materials selection. Often, a chrome-equivalent pipe will be specified (i.e., chrome plus a certain percentage of copper or another material.)
In modern supercritical plants, steam pressures range from 3,200 to 3,500 psi with temperatures maintaining about 1050°F to 1080°F (566°C to 582°C). In advanced and ultra-supercritical applications, the temperature is pushed even higher to nearly 1200°F (649°C). At this temperature, material families are typically shifted from a ferritic- to a nickel-based alloy because generally, after 1200°, the next jump is to at least 1300°F to 1350°F (704°C to 732°C), which requires the nickel-based materials.
Now that the temperatures in these plants have generally exceeded the practical limits of CS, P11 and P22 materials, engineers are specifying P91 and P92 in steam systems. For feedwater systems, the trend is up from Grades B and C to chrome equivalent (CrEQ) and P36 materials, especially in China and Asia.
Currently, for applications up to 800°F (427°C), CrEq carbon steel or P36 is used for applications where flow-accelerated corrosion is a concern. Between 800°F and 1025°F (427°C to 552°C), P11 and P22 materials are used. Above those temperatures, P91 or P92 material is used. Recently, the American Society of Mechanical Engineers (ASME) lowered the allowable stresses on P91, so the decision about when to shift from P91 and P92 has become more about economics than technical choices.
Alternate materials are a good solution because sometimes CrEQ materials are difficult to find.
P36 has become popular in India and China. It was developed in Germany specifically to reduce the rate of FAC in feedwater systems. P36 has a 35% reduction in wall thickness compared to carbon steel. The major disadvantage of this material is that there is no P36 casting equivalent for valves and other inline equipment. Because valves of another material cannot meet with the piping, it doesn’t make sense to buy a superior piping material because there would need to be a transition between the piping and valves. P36 has welding challenges similar to those that affect creep strength enhanced ferritic steel (CSEF).
CSEF is a family of high-alloy materials that contain between 9-12% of small amounts of various other materials including molybdenum, vanadium, niobium and varying additions of tungsten, cobalt, boron, nitrogen and nickel.
P91 is one such ferritic alloy steel. It has been in use for 25 years and is also called 9Cr-1Mo steel (based on its composition). Chemistry doesn’t define P91, however; it’s only half the equation for getting the strength these materials need to have. That strength also comes from grain structure caused by the heat-treating process—it is a tempered martensitic material.
The advantage of CSEF steels generally is that they are over twice as strong as P22 with half the wall thickness. This is very important because it reduces cost and increases flexibility.
The disadvantage of CSEF steels is that the strength is obtained through chemistry and heat treatments, so it is very easy to destroy the steel during the welding process. In a shop, that destruction can be controlled; but in the field, that’s very difficult. Welding requires strict controls to maintain martensitic grain structure, so preheat and post-weld treatment and interpass temperatures are tightly managed. P91 is listed as a P5B Weld Code by ASME, which means a post-weld heat treatment is mandatory. Weld heat-affected zones (HAZs) are the likely failure locations, and they can lead to type IV cracking. P91 is also susceptible to carbon migration and hydrogen embrittlement.
There are, however, other CSEF steels. In the last 15 years, materials used more often include P92, P911 and P122.
P92 was originally developed by the Japanese company Nippon Steel as NF616. P92 has about 30% reduction in wall thickness compared to P91 and includes tungsten in its composition. The addition of tungsten creates its own set of challenges regarding welding; but the main concern is the lack of a market for scraps (because of the tungsten).
P36 was developed by Vallourec & Mannesmann, which also developed a copper-nickel-molybdenum-alloyed carbon steel. These are standard materials in India and China for supercritical applications. P36 is approved under ASME Code Case 2353 and has been used in more than 30 plants since 1972.
The advantages of this material include having a 35% reduction in wall thickness over A106 Grade C as well as a high resistance to FAC. However, no casting equivalent currently exists so transition pieces or forged valves are needed.
Forged vs. Cast Steel Valves
The trend over the last few years is that forged valves are quoted as an alternative to cast steel valves, especially for high pressure applications. Forged valves have traditionally been used for small bore applications (2 inches and smaller and at the most, up to 8 inches). However, currently 18- to 24-inch sizes have been developed, but they’re not closed-form forged valves—they are valves machined out of a forged billet. These valves are showing up more and more, presenting unique challenges.
Some feel forged valves have better quality, and the wall thickness can be less. But both forged and cast valves can provide acceptable performance for most power applications. Forged valves are used more in streams where matching piping material to valve material is difficult. This is because code requirements such as ASME’s quality factor can cause design challenges with regards to transitions between pipe and valve materials and wall thicknesses.
With new materials, cast grades often lag in development, and forged valves provide more direct use of advanced materials. Designers are forced to specify a lower strength material, and transitions are required for differences in strengths in addition to dissimilar metal welding.
This has caused a delay in use of P92 and P36 materials in the U.S. market.
In the U.S., supercritical requirements are likely to remain the same, but new materials are continually developed. There are new ferritic steels and meta-stable austenitics coming into the market.
One is called SAVE12D (grade 93), a normalized and tempered steel that is 30% stronger in creep than P92 at 1200°F (649°C). This material has improved creep rupture ductility and is type IV-free for less-degradation welds. SAVE12D was approved in October of 2015 by ASME Section I Code Case 2839 and is a proposed designation of P93 for pipe, T93 for tube and F93 for forgings in American Society for Testing and Materials and ASME.
Another material on the horizon is HR6W (UNS N06674), which is a solution-annealed nickel alloy. This has stability of long-term creep rupture strength, superior creep rupture ductility and much better corrosion resistance than 18Cr-8Ni austenitic stainless steels. Also, it has microstructural phase stability at elevated temperatures, which contributes to superior stress relaxation and fatigue properties. Tests have shown better formability, wider available size range and better weldability than other nickel-based alloys.
What this means for the valve industry is that traditional mid- to high-alloy steels will be used for supercritical units operating below 1200°F (649°C). Higher fatigue cycle resistance will be required for fast-responding assets as renewables come to market, and nickel-based materials will be used for supercritical applications greater than 1200°F (649°C). This will most likely be limited in markets outside the U.S. and Europe.
Going forward, the valve industry’s great challenge when it comes to all this will be to make valves that are compatible with the new materials.