Materials in a reactor

After six months in thermal hydraulics, Hana Okoye's next rotation at Meridian Nuclear Engineering is with the materials and ageing group. Its lead, Dr. Malik Osei, meets her in a laboratory corridor lined with sectioned components: a slice of pressure tube with a faint band of oxide on its inner surface, a feeder elbow cut open to show where the wall has thinned, a fuel bundle end plate. "Thermal hydraulics tells you how hot things get," he says. "Materials tells you what that heat, the neutrons and the water do to the metal over forty or sixty years. At Darlington, every unit has just been refurbished with new fuel channels. Our job is to know, years ahead of time, how those new tubes will age."

Why materials limit reactor life

A power reactor is, among other things, a collection of metal and ceramic components that must keep their shape, strength and leak tightness in one of the harshest environments in engineering. The fuel must hold its fission products. The cladding or sheath must stay intact and coolable. The pressure boundary, whether a large steel vessel or many small zirconium alloy pressure tubes, must not fail by fracture or leak. Steam generator tubes separate radioactive primary coolant from the secondary side. Piping, pumps and valves must resist corrosion and fatigue for decades. Concrete and cables must survive radiation and heat. In practice, the economic life of many plants is set not by the reactor physics or the turbine but by how fast a few key components degrade and whether they can be inspected, repaired or replaced.

Materials engineering in a nuclear plant therefore has three jobs. The first is selection: choosing alloys and ceramics with the right properties at the start. The second is prediction: estimating how properties will change with neutron dose, temperature, stress and time, using test reactors, surveillance specimens and models. The third is management: inspecting, testing and acting before degradation reaches a limit. This course follows those three jobs from the atom to the ageing management program.

The reactor environment

Four features of the environment act together on reactor materials.

Each feature alone would be manageable; their combination is what makes reactor materials difficult. Irradiation speeds up creep that would be negligible at 300 °C without neutrons. Hydrogen from corrosion makes zirconium alloys more susceptible to cracking. Thermal cycling during start-ups and shutdowns adds fatigue to slowly growing flaws. A sound materials argument considers every mechanism that the combined environment can drive.

A tour of the components

In a CANDU reactor such as those at Darlington, the most important materials are distributed through hundreds of fuel channels rather than concentrated in one large vessel.

How properties are measured

Materials engineers rely on a few families of tests, each tied to a design question.

Irradiated specimens are tested in shielded hot cells, behind thick windows, using remote manipulators. The specimens themselves come from three sources: test reactors, where materials are exposed to high flux under controlled conditions; surveillance capsules placed inside operating reactors, which age alongside the real components; and samples removed from components during outages or refurbishment. Each source answers a different question, and the best predictions combine all three.

At Darlington, pressure tubes removed during refurbishment were a rare chance to measure the properties of tubes that had operated for about thirty years. Such data, together with periodic scrape samples and inspections of tubes in service, underpin the fitness-for-service evaluations that the CNSC expects for every operating unit.

Materials and safety requirements

Materials requirements appear in codes, standards and regulatory documents. Pressure-retaining components are designed and fabricated to the ASME Boiler and Pressure Vessel Code, whose Section III covers nuclear components and Section XI covers in-service inspection; both are used in Canada and the United States, alongside national standards. The licensing basis then requires that materials degradation be monitored and managed.

In Canada, CSA Group's N285 series of standards covers pressure-retaining systems and components in CANDU plants. CSA N285.4 sets requirements for periodic inspection of CANDU components, and CSA N285.8 sets technical requirements for in-service evaluation of zirconium alloy pressure tubes. The CNSC's regulatory document REGDOC-2.6.3, Aging Management, sets expectations for a licensee's ageing management program, which Lesson 12 discusses.

# Materials in a reactor: a simple screening table for one component.
# Each environment factor is scored 0-3; the products show which damage
# mechanisms dominate. Scores are illustrative teaching values.
components = {
    "fuel sheath (zirconium alloy)": {"T": 2, "flux": 3, "chem": 2, "stress": 2},
    "pressure tube (Zr-2.5Nb)":      {"T": 2, "flux": 3, "chem": 2, "stress": 3},
    "pressure vessel (low-alloy steel)": {"T": 2, "flux": 1, "chem": 1, "stress": 3},
    "steam generator tube (Alloy 690)":  {"T": 2, "flux": 0, "chem": 3, "stress": 2},
    "turbine-side piping (carbon steel)": {"T": 1, "flux": 0, "chem": 3, "stress": 2},
}
mechs = {
    "irradiation damage": ("flux", "T"),
    "creep and growth":   ("flux", "stress"),
    "corrosion":          ("chem", "T"),
    "cracking":           ("chem", "stress"),
}
for name, e in components.items():
    scores = {m: e[a] * e[b] for m, (a, b) in mechs.items()}
    top = max(scores, key=scores.get)
    row = ", ".join(f"{m} {v}" for m, v in scores.items())
    print(f"{name:36s} -> {row}; main concern: {top}")

fuel sheath (zirconium alloy)        -> irradiation damage 6, creep and growth 6, corrosion 4, cracking 4; main concern: irradiation damage
pressure tube (Zr-2.5Nb)             -> irradiation damage 6, creep and growth 9, corrosion 4, cracking 6; main concern: creep and growth
pressure vessel (low-alloy steel)    -> irradiation damage 2, creep and growth 3, corrosion 2, cracking 3; main concern: creep and growth
steam generator tube (Alloy 690)     -> irradiation damage 0, creep and growth 0, corrosion 6, cracking 6; main concern: corrosion
turbine-side piping (carbon steel)   -> irradiation damage 0, creep and growth 0, corrosion 3, cracking 6; main concern: cracking

Thinking in mechanisms

Malik's first lesson for Hana is to think in mechanisms rather than in symptoms. A crack found in an inspection is a symptom. The mechanism might be fatigue, stress corrosion cracking, hydride cracking or a fabrication flaw; each has different causes, different growth rates and different remedies. The same is true of a change in dimension, which might be creep, growth or swelling. The rest of this course works through the main mechanisms in turn: radiation damage and its effects on strength and toughness, the behaviour of fuel, zirconium alloys and steels, creep and growth, corrosion, moderator materials, accident-tolerant fuels and materials for advanced reactors, and finally how all of this is tied together in an ageing management program.

Why so few materials are used in the core

The choice of materials inside the core is narrower than in almost any other industry, because every material placed near the fuel also affects the neutrons. A structural metal in the core must absorb few neutrons, keep its properties under irradiation, resist corrosion by the coolant and not become too radioactive for maintenance. Zirconium is used for cladding and pressure tubes mainly because its thermal neutron absorption cross section is very small, about 0.18 barn, compared with about 2.6 barns for iron and nearly 4.5 barns for nickel. Natural zirconium ore contains about 2 percent hafnium, a strong absorber, so reactor-grade zirconium must be specially purified to remove it.

Outside the core, the constraints change. Pressure vessels and piping are made of steel because steel is strong, tough, weldable and well understood, and because the flux at the vessel wall is low enough that absorption hardly matters. Steam generator tubes need corrosion resistance and good heat transfer above all, which leads to nickel-chromium-iron alloys. In each region, the selection is a compromise among nuclear, mechanical, chemical and economic properties.

The point about induced activity deserves emphasis. Cobalt-59 in steels and hard-facing alloys absorbs neutrons to form cobalt-60, a strong gamma emitter with a half-life of about 5.3 years. Corrosion releases small amounts of cobalt into the coolant, which then deposits on piping outside the core and raises radiation fields where people work. Specifications for reactor steels therefore limit cobalt content, and plants have replaced cobalt-based hard-facing in valves with cobalt-free alternatives where they can. Materials choices made at the design stage affect worker dose for the life of the plant.

Learning from operating experience

Much of what is known about reactor materials was learned the hard way, from degradation found in operating plants. Stress corrosion cracking of stainless steel piping in boiling water reactors in the 1970s and 1980s led to changes in alloys, welding practice and water chemistry. Cracking of Alloy 600 steam generator tubes and nozzle welds led to the use of Alloy 690. In CANDU reactors, early pressure tubes of Zircaloy-2 and a few tubes that cracked from hydride blisters led to the use of zirconium–2.5 % niobium and to the garter spring inspection and repositioning programs. Each event fed back into codes, standards and inspection programs, and operating experience exchanged among plants and countries remains one of the main inputs to every ageing management program.