Structural integrity challenges associated with long term operation of civil nuclear power plant

Abstract

The UK is entering a new era for civil nuclear power which involves extending the operating life of existing plant, building new plant and designing plant for future fission reactors (Gen IV). The latter two will be required to operate for periods which present challenges for the present materials and methodologies. This paper reviews these potential challenges in the context of the current understanding of materials performance and existing assessment methodologies. In addition, it provides an example of an unexpected creep-related challenge arising from long term operation of Type 316H boiler tubes.

Keywords

Nuclear boilers stainless steel creep impurity elements precipitation service life

1. Introduction

From the earliest days of civil nuclear electrical power generation in the UK, the structural integrity of key components has been recognised to be an essential factor related not just to safety but also to economic operation. The main structural integrity inputs when a particular operational problem arises are inspection and monitoring, together with diagnosis and assessment (Fig. 1). The UK is now entering a new era for the civil nuclear power industry. In addition to extending the operating life of the UK fleet of advanced gas-cooled reactors (AGRs) to about forty years, both new build plant and the designs of plant for future fission reactors, classed as Generation IV (Gen IV), will be required to operate for periods of sixty years plus. This presents challenges for present materials and assessment methodologies to achieve this ambition. An essential input for demonstrating the necessary assurance is a structural integrity assessment. To achieve the defined service life, the design engineer is required to create a structure or component using analytical tools and data most appropriate for the application. Subsequently, the operator has to ensure that throughout the service life it is secure against the design intent [1–4]. For these, it is essential to have a knowledge of the mechanical and physical properties of the materials, both for construction and performance over the operating life.

Fig. 1.

Schematic diagram showing the main inputs when diagnosing structural integrity problems and the associated output options.

2. Structural integrity requirements

As pointed out previously by Dowling and Flewitt [5], design is focused on expectation, whereas safe continued operation is more concerned with reality. As a consequence, for plant design and through-life management to be secure we need to have a detailed understanding of how materials behave over a wide range of length-scale. As summarised in Fig. 2, this should span the atomic behaviour (nano-length-scale) which influences and controls the evolution of the microstructure of material and the micro-mechanical processes when subject to stress, temperature and environment. Finally, it is these processes that influence the macro-scale mechanical properties which, in turn, are the necessary inputs to the assessment methodologies adopted when assessing structures and components in the various service environments. Ideally, rather than making measurements at each length-scale we require predictive models that can be incorporated over the scale-range to predict the appropriate macro-scale properties for input to the structural integrity assessments. We are far removed from this capability, but this should not deter the intent.

Fig. 2.

For plant design and through-life management there is a need to understand: (a) atomic processes; (b) evolution of microstructure; (c) micromechanical processes; and (d) macroscopic mechanical properties; which input to (e) the final application.

A major principle in nuclear safety is ‘Defence in Depth’ which, wherever possible, is sought through diversity, redundancy and segregation to ensure and demonstrate a robust and fault tolerant design. However, for some structures and components it is not possible to satisfy such requirements by physical means. In cases where the failure of a single component would result in unacceptable consequences for public safety, ‘Incredibility of Failure’ is sought, whilst retaining the principle of ‘Defence in Depth’ through application of appropriate experience, testing, analysis and monitoring.

Since a safety case is a formal statement of comprehensive and systematic safety assessments, it is a requirement to demonstrate that the risks to workers and the public are sufficiently low. Where multiple physical barriers to failure do not exist, it is necessary to introduce the alternative of ‘Conceptual Defence in Depth’ arguments. When asked to address this challenge, the UK Technical Advisory Group for Structural Integrity (TAGSI) [6,7] argued that four conceptually different legs of a safety case may be defined, namely: Leg 1: Interpolation/Extrapolation of Experience - ‘having done it before’; Leg 2: Functional Testing - ‘having demonstrated functionality by representative testing’; Leg 3: Failure Analysis - ‘fitness for purpose using current best scientific understanding’; and Leg 4: Forewarning of Failure - ‘the promise of future action on gaining new information’. These four legs were perceived to form the basis of any argument for inclusion in a safety case. For example, in the case of steel pressure vessels these legs become: Leg 1 - High reliability derived from historic experience of good design and construction as typically expressed in design codes; Leg 2 - An overpressure or proof test; Leg 3 - An assessment of failure modes, for example via fracture mechanics; and Leg 4 - Plant monitoring or in-service inspection. Therefore, a knowledge of materials ageing and degradation characteristics is fundamental to Legs 3 and 4. As pointed out by Burdekin [7], the risk of creep failure is dominated by selection of material for the required operating temperatures and control of the operating conditions to ensure that both the loading and temperature are maintained within the prescribed design limits. To achieve this, Leg 1 has to be strong and Leg 3 may be required to contribute to the detailed stress analysis and failure considerations. However, for these arguments Leg 2 contributes little, whereas Leg 4 provides the option of monitoring for leakage. Finally, an engineering judgement that the plant is fit for duty is made on the basis of the evaluation of all of the legs.

Fracture mechanics as a basis of assessment of the integrity of engineering structures moved forward with the publication of ASTM STP381 ‘Fracture Toughness Testing and its Applications’ [8] and the associated concepts of crack tip opening displacement. This led to the R6 procedure that was developed and first published in 1976 for assessing defects in electrical power generating plant structures and components entitled ‘Assessment of the Integrity of Structures Containing Defects’ [9–12]. It was based on the two principal failure criteria proposed by Dowling and Townley [13]: (i) linear elastic fracture mechanics; and (ii) plastic collapse. These two criteria were adopted to define a failure assessment diagram (Fig. 3) that is equivalent to a J-integral analysis combined with plastic collapse. Here, the two parameters K_r and L_r are defined as: $\begin{eqnarray}\displaystyle \mbox{}&\mbox{}& K_{r}=(K_{I}^{P}+VK_{I}^{S})/K & & \displaystyle\end{eqnarray}$ (1) $\begin{eqnarray}\displaystyle \mbox{}&\mbox{}& L_{r}=P/P_{L}(a,{\sigma}) & & \displaystyle\end{eqnarray}$ (2) where $K_{I}^{P}$ , $K_{I}^{S}$ are stress intensity factors for primary and secondary stresses, V allows for plasticity interaction between primary and secondary stress, K is fracture toughness, P the applied load and P_L is the plastic collapse load. Over the intervening years, this procedure has been continually subject to update and revision. Indeed, it provides an excellent basis for avoiding failure of a structure or component in service [5]. R6 is now in Revision 4 [12] and is maintained by EDF Energy Nuclear Generation in collaboration with a number of other organisations.

Fig. 3.

A typical R6 failure assessment diagram showing an assessment point within the safe region, together with the trajectories of this point when fracture toughness K_lc load and crack size are varied. Axes K_r and L_r are defined in the text.

To accommodate operating at higher temperatures, the R5 document was developed in the early 1990s, entitled ‘Assessment Procedure for High Temperature Response of Structures’ [14–17]. The procedures described are relevant to creep, fatigue and creep-fatigue loadings and allow assessment of components that are defect-free and contain crack-like flaws. The objective of R5 is to provide a comprehensive assessment procedure for high temperature operation. It augments, where appropriate, the ASME [18] and French RCC-MR [19] design codes, and the materials data associated with the R5 procedure are contained in a complementary document [20]. Both procedures provide an excellent basis for avoiding failure of components and structures during service.

3. Challenges for long term operation

In the immediate term in the UK there is: (i) life extension to 40 years for the AGRs and the one pressurised water reactor (PWR) power station; and (ii) the new build programme of various plant, including two PWR designs, the EPR and the Westinghouse AP1000, together with an advance boiling water reactor (ABWR). In the latter case, there is a requirement to operate for design life periods of sixty years. Although the reactor designs are based upon mature technology, these plants operate at relatively low temperatures so that there is the potential for both materials and methodology challenges to achieve the declared design life. One challenge arises from environmentally-assisted cracking.

In the case of environmentally-assisted cracking, and the related issues of corrosion fatigue, creep deformation and failure, the assessment procedures are based upon constitutive equations. The drawbacks of such approaches are: (i) the difficulty of transferring the models from one test condition to another; and (ii) the limitation of predicting material behaviour beyond the range of the test data. As a consequence, there is the need to develop improved mechanistic models. Unfortunately, for environmentally-assisted cracking and creep the models to date provide only good prediction for nominally pure metals and simple solid solution alloys. Therefore, for improvised assessment methodologies the development of physically based models is required and remains a significant challenge.

As described elsewhere [21,22], it is clear that since the turn of the millennium most advanced nuclear systems have been considered under the Gen IV initiative. Within this framework, six systems have been selected [22]:

•
Gas-cooled fast reactor (GFR)
•
Very high temperature reactor (VHTR)
•
Sodium-cooled fast reactor (SFR)
•
Lead-cooled fast reactor (LFR)
•
Molten salt reactor (MSR)
•
Super critical water-cooled reactor (SCWR)

The aim is to operate at a high temperature to: (i) achieve good thermal efficiency; and (ii) support the production of hydrogen from water by thermo-chemical processes. The metallic alloys suitable for such high temperature operation are limited to, in general, austenitic stainless steels and nickel-based alloys. Hence, the main challenge arises from creep-related problems. In the UK, assessment methodologies, such as R6 ‘Assessment of Integrity of Structures Containing Defects’ [12] and R5 ‘Assessment Procedure for High Temperature Response Structures’ [17], are adopted. Although both R6 and R5 procedures provide an excellent basis for failure avoidance during service of components and structures, they are now mature. Developments are still underway but there must be a question as to their appropriateness for application to the new era of nuclear power generation plant where periods of operation of ≥60 years are required. Hence, it is appropriate to consider whether procedures based upon improved physical understanding of the mechanisms controlling failure, both at low and higher temperatures, should be developed. Certainly, assessments have to take into account the longer-term time-dependent changes in material properties arising from the hostile environments, which can be gaseous, aqueous, irradiation and temperature under service loads.
3.1. Environmentally-assisted cracking

Environmentally-assisted cracking remains a potential threat to the long term operation, of at least sixty years, for nuclear power plant that operates at lower temperatures, such as the pressurised water and boiling water reactors. Hence, there is a need to consider this challenge for the new build generation of reactors as well as future designs associated with Gen IV [23–25]. In general, a range of materials have been and will be used for the construction of nuclear power generating plants that fulfil the design intent. However, for extended periods of operation there is less experience on which to base such evaluations. Environmentally-assisted cracking spans several forms, including intergranular attack, stress corrosion cracking and corrosion fatigue. It is widely recognised and, indeed, well established that there are three prerequisites for environmentally-assisted cracking to occur (Fig. 4): (i) a susceptible material; (ii) an appropriate environment; and (iii) a tensile stress, either applied or residual. In general, stress corrosion cracking occurs under steady or slowly varying loads whereas corrosion fatigue is associated with cyclic loads. Often, stress corrosion cracks have been observed to arise from a three-stage process. In the first stage there is degradation of the smooth surface as a precursor to the cracking. As shown by Doig and Flewitt [26,27], this stage can involve localised corrosion at preferential sites leading to the development of occluded cells that fill with corrosion products. As a consequence, a local pit is formed. The next stage combines the local stresses-strains associated with these local pits to initiate a stress corrosion crack. Turnbull [28], using x-ray computed tomography, showed that the transition from a pit to a crack does not necessarily occur at the base of such a pit. Rather, it can occur at the side of a pit where finite element analysis revealed localised dynamic strain to be sufficient to initiate a stress corrosion crack. The final stage is crack propagation, where the growth rate is controlled by the material microstructure and composition along the crack path, the applied stress intensity and solution environment. The crack can follow either a transgranular or intergranular path at that stage. However, as shown by the Venn diagram (Fig. 4), the complexity of these many contributions make it difficult to ensure stress corrosion cracking will not be encountered in plant components over long term operation, because of the challenges of potentially changing material properties, environment and service load.

Fig. 4.

A Venn diagram showing the main contributions to environmentally-assisted cracking.

Environmentally-assisted cracking in both BWR and PWR plant has led to stress corrosion cracking and corrosion fatigue, resulting in coolant leakage rather than catastrophic failure [26]. In many cases, in-service inspection has identified such cracking occurring in both PWR and BWR materials, including:

•

Ferritic steels - Low-alloy steels such as A533B and A508-III used for main pressure vessels, carbon steels used for PWR steam generator tubesheets, piping in some BWRs and various other piping applications, and high strength quench and tempered steels used for bolting applications.

•

Stainless steels - Type 300 series austenitic stainless steels used widely in both BWRs and PWRs for piping, pump and valve bodies and a variety of other applications. Martensitic stainless steels are employed where higher strength is required, such as in valve stems and some fastener applications. Precipitation-hardened alloys, such as A-286 and 17-4PH, are also used for high strength applications.

•

Nickel-based alloys - Alloy 600 has been extensively used in PWRs, for steam generator tubing and other applications, including control rod drive penetrations, bottom head penetrations and steam generator divider plates. The weld metal alloys 182, 82 and, in Japan, 132. Alloy 690 and weld metal alloys 52, 152 and variants are being widely used for replacement and new build due to relatively poor service experience of Alloy 600. Precipitation-hardened alloys X-750 and 718 are employed for higher strength requirements, such as internal bolting, fasteners and springs.

There is clearly a requirement to develop mechanistically-based models to predict the service life of components with improved confidence to achieve not less than sixty years of operation. In addition, there is a need for significantly improved assessment methodologies and codes. In the case of the R6 failure avoidance methodology there is little more than simple guidance provided to accommodate this complex problem. Moreover, although the ASME Boiler and Pressure Vessel Code provides rules for the design of Class 1 components of nuclear power plants and Appendix I to Section III of the Code specifies fatigue design curves for applicable structural materials, the effects of light water reactor (LWR) coolant environments are not explicitly addressed [29]. The existing fatigue strain versus life (𝜖-N) data illustrate potentially significant effects of LWR coolant environments on the fatigue resistance of pressure vessel and piping steels. Under certain environmental and loading conditions, fatigue life in water relative to that in air can be a factor of ∼12 lower for austenitic stainless steels, ∼3 lower for Ni–Cr–Fe alloys and ∼17 lower for carbon and low-alloy steels. Subsequently, NUREG/CR609 provided guidance on fatigue of piping and pressure vessel assessments in LWR environments [30]. This has been further updated to address concerns related to: (i) the constants in the environmental fatigue correction factor (F_en) expressions that result in values of about two even when either the strain rate is very high or the temperature is very low; (ii) the temperature dependence of F_en for carbon and low-alloy steels; and (iii) the dependence of F_en on water chemistry for austenitic stainless steels. The F_en methodology was validated by comparing the results of five different experimental data sets obtained from fatigue tests that simulated plant conditions. However, there is clearly much to do to provide more realistic approaches to accommodate stress corrosion and corrosion fatigue cracking to meet the ≥60 year life time challenge.

3.2. Creep

A candidate material for the design of GFR and VHTR main structural components is the austenitic stainless steel AISI Type 316L(N). Here, the major alloying elements are 18Cr, 12Ni and 2Mo (all wt%). The L denotes low carbon ≤0.03 wt% and the N denotes additions of nitrogen in the range 0.06 to 0.08 wt%. In this context, the UK has unique experience of operating the AGR plant at temperatures of the order of 550 °C over extended periods and with a related material, ISI Type 316H austenitic stainless steel.

There are two significant interactive contributions to the creep lifetime of engineering materials, i.e. creep deformation and creep fracture [31]. Materials may deform by several different mechanisms when subjected to an applied stress at high temperature, and it is convenient to present these mechanisms in the form of a deformation mechanism map [32,33]. Similarly, materials may fracture by one of several possible mechanisms, and these can be described by a fracture mechanism map [34,35]. More importantly, over the operational service life there is a potential to change the initial microstructure of a material by thermal ageing which can affect the controlling deformation and fracture mechanisms.

In the specific case of Type 316 austenitic stainless steel, a typical deformation mechanism map for the steady state creep response has been established by Frost and Ashby [33]. The deformation mechanism map was generated by fitting generic models to creep data. In general, power law creep, dominated by dislocation movement, and diffusional creep are the two main high temperature deformation mechanisms. The former operates at a relatively high stress and the latter at a relatively low stress. Two widely accepted physically-based models that describe diffusional creep deformation were proposed by Nabarro-Herring [36,37] and Coble [38]. Data to support the diffusional creep mechanism are limited, and most designs are assumed to fall within the regime where dislocation mechanisms dominate the material response. Many different forms of constitutive equations have been proposed to describe dislocation-dominated creep deformation [39–45]. These include the power-law relationships employed by Frost and Ashby [32] and exponential and hyperbolic sine relationships [44]. However, to date no unified mechanistic models have been proposed and most of the equations simply provide a functional form of constitutive model that can be fitted to data. Thus, the resulting models can be used to predict creep behaviour only within the bounds of test data [46]. There are at least two obvious drawbacks associated with the development of constitutive equations in this way: (i) the difficulty of transferring the model from one tested material to another; and (ii) the limitation of predicting material behaviour beyond the test data range, i.e. life evaluation. Hence, it does not avoid the requirement for expensive creep test programmes to develop long term creep lifetime predictions.

Microstructural changes can result from prolonged exposures at temperatures up to 650 °C. The kinetics and type of precipitates formed depend upon several factors, including: (i) the specific composition of the material; (ii) the microstructure arising from the thermo-mechanical history when entering service; and (iii) the state and magnitude of the service stresses. Different secondary phases evolve in Type 316H austenitic stainless steel during long term ageing and service [47], including 𝛼-ferrite, carbides and intermetallic phases [47–55]. Carbides $\mathrm{M}_{23}\mathrm{C}_6$ and $\mathrm{M}_6\mathrm{C}$ , where M is Fe, Cr, Ni or Mo, have been identified in Type 316H austenitic stainless steel, but other types, such as $\mathrm{M}_7\mathrm{C}_3$ or MC, may also evolve, depending on the specific composition [47 ]. Such precipitates can lead to a change in creep deformation rate by pinning dislocations [51 ], but the presence of specific precipitates, such as 𝛼∕𝛿-ferrite, at grain boundaries can promote creep cavitation. It is also recognised that intermetallic phases, such as sigma and chi, degrade creep life. All of these precipitates have the potential to form over the service life by thermal ageing. Phase diagrams provide a useful guide to the phases present in stainless steel in equilibrium. However, the situation becomes complex in application to plant in service leading to the need to adopt multi-component thermodynamic calculations. At present, their usage to predict the phases present as a result of long term service is, at best, an estimation, and this can lead to unexpected issues for assessing the structural integrity of long term service plant. It is in this context that the UK has unique experience since the AGR power plant has operated at high temperature (∼550 °C) for periods in excess of 30 years. Martinez-Ubeda [56] and Martinez-Ubeda et al. [51,57] have described one such unexpected concern encountered in AISI Type 316H reheater boiler tubes.

Fig. 5.

Electron backscatter diffraction map of Tube B showing $\mathrm{M}_{23}\mathrm{C}_6$ carbide and ferrite precipitates distributed in the austenitic matrix [57].

Fig. 6.

(a) An area 150 μ $\mathrm{m}$ ×150 μ $\mathrm{m}$ mapped by high speed atomic force microscopy showing the parent austenite for Tube A. (b) Selected frame showing intergranular precipitates (light contrast) and associated creep cavities (dark contrast) [51].

Austenitic stainless steel Type 316H is used in the AGRs as a material for boiler superheaters due to its good resistance to oxidation and high temperature mechanical properties. The nominal composition of austenitic stainless steel is defined by AISI standards to be within a specific range of composition and there is a measure of acceptable variability to accommodate cast-to-cast differences. Three reheater bifurcation tubes produced by different manufacturers, A, B and C, but to the same specification for Type 316H stainless steel, had operated at 505 °C for extended periods (Table 1). In Tube A, there were creep-induced cracks, whereas in the other two (B and C) there was some distributed cavitation, but no cracking. Table 2 shows the compositions, where it is noteworthy that Tube B and Tube C contain higher concentrations of silicon, whereas in Tube A the nickel concentration is lower. All tubes contain impurity phosphorus to a concentration of ∼0.02 wt% to ∼0.03 wt%.

Fig. 7.

High resolution transmission electron micrographs showing atomic resolution contrast for: (a) austenite; (b) $\mathrm{M}_{23}\mathrm{C}_6$ carbide; and (c) G-phase in Type 316H austenitic stainless steel.

Fig. 8.

Energy dispersive x-ray analysis composition distributions: (a) across the $\mathrm{M}_{23}\mathrm{C}_6$ -austenite interphase boundary in Tube A; and (b) across a G-phase precipitate (silicon-enriched) in Tube C (mid-section to interphase boundary).

All three tubes had similar microstructures compared to the one shown in Fig. 5. This is an electron backscatter diffraction (EBSD) image of Tube A, revealing austenite grains, typically ∼40 μ $\mathrm{m}$ grain size (mean linear intercept) with both $\mathrm{M}_{23}\mathrm{C}_6$ carbide precipitates and ferrite precipitates at the grain boundaries. Figure 6(a) shows a corresponding high speed atomic force microscopy (HSAFM) mapped montage of the microstructure for Tube A where, following vibropolishing, the carbide precipitates are shown in relief at the grain boundaries. Figure 6(b) provides detail of an austenite boundary, where the light contrast corresponds to the carbide features, extending typically to ∼200 nm above the austenite surface. The local dark areas are below the surface and are consistent with grain boundary creep cavities. However, detailed high-resolution transmission electron microscopy revealed that for Tube B and Tube C, in addition to the grain boundary $\mathrm{M}_{23}\mathrm{C}_6$ carbide and ferrite precipitates, these higher concentration silicon materials (Table 2) also contained intragranular silicon-enriched G-phase precipitates with an fcc structure and lattice parameter of ∼1.12 nm. Figures 7(a)–(c) show typical atomic resolution images of: (a) the austenite matrix; (b) an associated grain boundary $\mathrm{M}_{23}\mathrm{C}_6$ carbide precipitate; and (c) an intragranular G-phase precipitate. When combined with energy dispersive x-ray analysis, this revealed segregation of impurity phosphorus to the carbide-austenite interphase boundary to ∼12 wt% for Tube A (Fig. 8(a)). By comparison, no phosphorus segregation was identified at either the austenite grain boundaries or the carbide-austenite interphase boundaries in the cases of Tube B and Tube C. However, in these latter tube materials containing G-phase, there was partitioning of phosphorus to the precipitates (Fig. 8(b)). It was concluded that this redistribution of impurity phosphorus to the silicon-enriched G-phase precipitates precluded segregation to the interphase boundaries between the carbide and austenite.

Table 1

Three tubes (A, B, C) from different manufacturers showing service conditions and presence of grain boundary cracking

Sample	Service time	Temperature	Cracked?
	(h)	(°C)
A	160000	505	Yes
			(grain boundary cavitation)
B	145000	505	No
C	300000	505	No

Table 2

Composition of tubes A, B and C within the specification for AISI Type 316H austenitic stainless steel

Sample	Average composition (wt%)
	Si	Fe	Ni	Cr	Mn	Mo	C	P
A	0.45	Bal.	11.48	16.21	1.43	2.34	0.05	0.02
B	0.64	Bal.	13.59	15.80	1.74	2.14	0.06	0.02
C	0.59	Bal.	14.07	16.49	1.19	2.04	0.07	0.02

It is recognised that the critical radius, r, for creep cavity nucleation can be expressed as [58]: $\begin{eqnarray}\displaystyle r=2{\gamma}/{\sigma} & & \displaystyle\end{eqnarray}$ (3) where 𝛾 is the interfacial energy and 𝜎 is the local tensile stress normal to a grain or interphase boundary. Any decrease in the interfacial energy, such as arising from impurity segregation, leads to a decrease in the critical cavity radius and, therefore, facilitates the nucleation of cavities. A key outcome from the work of Martinez-Ubeda et al. [51] is that creep cavities were associated with the grain boundary carbide precipitates in Tube A where there was 12 wt% enrichment in phosphorus. Similar enrichment was not present in Tube B or Tube C. Hence, there was enhanced cavity nucleation in Tube A leading to the observed crack formation at the boundary regions. Certainly, the unexpected formation of G-phase arising from long term service ageing had a critical impact on the ability of the phosphorus to segregate to the interphase boundaries and influence on long term integrity of the boiler tubes.

4. Concluding remarks

It is clear that the era where sixty years of operation of nuclear plant is expected presents several major challenges to safe and economic operation. In particular, there is a need to improve models that are physically-based so that the evaluation of service life can be set in a secure framework. In addition, the structural integrity assessment methodologies at low and high temperature, such as R6 and R5, may be insufficient and there is a need to review these and potentially move to a mechanistic basis if the sixty year challenge is to be successfully achieved.

Footnotes

Acknowledgements

PEJF would like to acknowledge the many colleagues he has worked with over a number of years and the discussions with them that have led to this paper.

Conflict of interest

None to report.

References

Flewitt

P.E.J.

, Mechanical Behaviour of Materials, in: Bakker

(ed), Delft University Press, 1995.

Knott

J.F.

, Two steps from disaster - The science and engineering of structural integrity, in: The Royal Society and The Royal Academy of Engineering 1999 Lecture, published by The Royal Society and The Royal Academy of Engineering1999.

Goranson

U.G.

, International Journal of Fatigue16 (1994), 43.

Harris

D.O.

Wells

C.H.

Rau

S.A.

and Dedhia

D.D.

, International Journal of Pressure Vessels and Piping59 (1994), 175.

Dowling

A.R.

and Flewitt

P.E.J.

, Methods for the Assessment of Structural Integrity of Components and Structures, in: Lidbury

and Hirsch

P.B. Maney

(eds) 2003.

Bullough

Burdekin

F.M.

Chapman

O.J.V.

Green

V.R.

Lidbury

D.P.G.

Swingler

J.N.

and Wilson

, International Journal of Pressure Vessels and Piping78 (2001), 539.

Burdekin

F.M.

, Some further thoughts on incredibility of failure, TAGSI Paper 108 (1996).

Special Technical Publication 381, American Society for the Testing and Materials (1965).

Harrison

R.P.

Loosemore

and Milne

, Assessment of the integrity of structures containing defects, Central Electricity Generating Board, Report R/H/R6 (1976).

10.

Harrison

R.P.

Loosemore

and Milne

, Assessment of the integrity of structures containing defects, Central Electricity Generating Board, Report R/H/R6-Revision 1 (1977).

11.

Harrison

R.P.

Loosemore

Milne

and Dowling

A.R.

, Assessment of the integrity of structures containing defects, Central Electricity Generating Board, Report R/H/R6-Revision 2 (1980).

12.

Milne

Ainsworth

R.A.

Dowling

A.R.

and Stewart

A.T.

, International Journal of Pressure Vessels and Piping32 (1988), 3.

13.

Dowling

A.R.

and Townley

C.H.A.

, International Journal of Pressure Vessels and Piping3 (1975), 77.

14.

R5 Assessment procedure for the high temperature response of structures, Issue 1, Nuclear Electric (1990).

15.

R5 Assessment procedure for the high temperature response of structures, Issue 2, Nuclear Electric (1992).

16.

R5 Assessment procedure for the high temperature response of structures, Issue 3, British Energy Generation (2003).

17.

R5 Assessment procedure for the high temperature response of structures, Issue 3, Revision 001, EDF Energy Nuclear Generation (2012).

18.

ASME Boiler and pressure vessel code, Section III, Division 1 - Subsection NH, Class 1 components in elevated temperature service, ASME (2010).

19.

RCC-MRx: Design and construction rules for mechanical components of nuclear installations, AFCEN (2007).

20.

Hamm

C.D.

, (ed) AGR materials data handbook, R66, revision 009, EDF Energy Nuclear Generation (2011).

21.

Status report on structural materials for advanced nuclear systems, OECD Nuclear Energy Agency (2013).

22.

Abram

, A technology road map for Generation-IV nuclear systems, US Department of Energy (2002).

23.

Speidel

M.O.

, Mechanical Behaviour of Materials, in: Miller

K.T.

and Smith

R.F.

(eds), Pergamon, 1979.

24.

Tice

D.R.

, Corrosion Science25 (1985), 705.

25.

Tice

D.R.

, Structural Integrity of Nuclear Power Plant, in: Knott

J.F.

and Tomkins

B.D.

(eds), EMAS Publishing, 2013.

26.

Doig

and Flewitt

P.E.J.

, Metallurgical Transactions A14 (1983), 978.

27.

Doig

and Flewitt

P.E.J.

, Mechanisms of Environment Sensitive Cracking in Materials, in: Swann

P.R.

Ford

F.P.

and Westwood

A.R.C.

(eds), The Metals Society, 1977.

28.

Turnbull

, Proceedings of the Royal Society A470 (2014), 20140254.

29.

Regulatory Guide 1.207, Guidelines for evaluating fatigue analyses incorporating the life reduction of metal components due to the effects of the light-water reactor environment for new reactors, US Nuclear Regulatory Commission (2007).

30.

Chapra

and Stevens

G.L.

, Effect of LWR coolant environments on the fatigue life of reactor materials (NUREG/CR-6909, Revision 1) - Draft, US Nuclear Regulatory Commission (2014).

31.

Evans

R.W.

and Wilshire

, Creep of Metals and Alloys, The Institute of Metals, 1985.

32.

Ashby

M.F.

, Acta Metallurgica20 (1972), 887.

33.

Frost

H.J.

and Ashby

M.F.

, Deformation-mechanisms Maps, Pergamon, 1982.

34.

Fields

R.J.

Weerasooriya

and Ashby

M.F.

, Metallurgical Transactions A11 (1980), 333.

35.

Miller

D.A.

and Langdon

T.G.

, Metallurgical Transactions A10 (1979), 1635.

36.

Nabarro

F.R.N.

, Report of a Conference on the Strength of Solids, The Physical Society, 1948.

37.

Herring

, Journal of Applied Physics21 (1950), 437.

38.

Coble

R.L.

, Journal of Applied Physics34 (1963), 1679.

39.

Kassner

M.E.

, Fundamentals of Creep in Metals and Alloys, Elsevier, 2009.

40.

Sherby

O.D.

and Burke

P.M.

, Progress in Materials Science13 (1967), 323.

41.

Takeuchi

and Argon

A.S.

, Acta Metallurgica24 (1976), 883.

42.

Orlova

and Cadek

, Materials Science and Engineering77 (1986), 1.

43.

Kassner

M.E.

and Perez-Prado

M.T.

, Progress in Materials Science45 (2000), 1.

44.

Gittus

, Creep, Viscoelasticity and Creep Fracture in Solids, Applied Science Publisher, 1975.

45.

Takeuchi

and Argon

A.S.

, Journal of Materials Science11 (1976), 1542.

46.

Blum

, Materials Science and Engineering A319 (2001), 8.

47.

Padilha

A.F.

and Rios

P.R.

, ISIJ International42 (2002), 325.

48.

K.H.

Shek

C.H.

and Lai

J.K.L.

, Materials Science and Engineering: R: Reports65 (2009), 39.

49.

Lai

J.K.L.

, Materials Science and Engineering61 (1983), 101.

50.

Weiss

R.S.

and Stickler

, Metallurgical Transactions3 (1972), 851.

51.

Martinez-Ubeda

A.I.

Griffiths

Oliver

Younes

C.M.

Scott

T.B.

and Flewitt

P.E.J.

, ASME Pressure Vessels and Piping Conference2016.

52.

Chen

Flewitt

P.E.J.

Cocks

A.C.F.

and Smith

D.J.

, International Materials Reviews60 (2015), 1.

53.

Mohamed

F.A.

and Langdon

T.G.

, Acta Metallurgica22 (1974), 779.

54.

Dyson

B.F.

, Materials Science and Technology25 (2009), 213.

55.

Dye

and Reed

R.C.

, Acta Materialia56 (2008), 1657.

56.

Martinez-Ubeda

A.I.

, PhD thesis, University of Bristol (2017).

57.

Martinez-Ubeda

A.I.

Griffiths

Karunaratne

M.S.A.

Flewitt

P.E.J.

Younes

C.M.

and Scott

T.B.

, Procedia Structural Integrity2 (2016), 958.

58.

Dyson

B.F.

, Scripta Metallurgica17(1983) 31.