Search

The EU project SmartResilience aims to develop methods for practical assessment of the resilience of smart critical infrastructures (SCIs) in urban environments. The paper will describe an approach applied in the case of the energy supply system of the city of Helsinki, Finland. One selected scenario involves a cold winter period challenging the supply and performance of the city district heating network that provides heating for most buildings of the city. A fire in an underground fuel storage initiates an escalating chain of events to reduce production capacity first in the adjacent plant and then in another heating plant of the network. Simultaneous extended period of high demand during the cold spell will trigger regional cooling in the district heating network, so that elderly and other vulnerable citizens may need to be relocated for safeguarding when they cannot rely on their own back-up alternatives. The scenario implies cascading involvement of city infrastructures and actors beyond the energy supply system. The details will affect the essential aspects, phases and indicators of functionality and resilience, and the effort to improve or optimize resilience. The tools developed for the project have helped to describe, visualize and quantify resilience for these purposes. The results and implications are discussed from the point of view of other plausible threats from extreme weather events or interdependencies in the evolving regional and European energy systems.

Risk-based inspection and maintenance can help to achieve safe and economically viable plant operation by justifying and prioritizing actions, choices and schedules according to the associated risk. To manage the risk, the implemented inspections and maintenance actions aim to limit the probability of failure and its uncertainty. Failure statistics of similar components operating in similar environments may provide fair estimates of the probability of failure, but biased or lacking comparable statistics may require adjustment through additional measurements or expert information. In parallel, inspection and monitoring data can more directly help in assessing the evolution of probability of failure. Presented example is a tubular heat exchanger, with inspection and failure criteria considered for a full life cycle that included multiple inspections by the same inspecting party and equipment. Uneven distribution of in-service wall loss without changes in mechanisms or local loss rates nevertheless required a change of inspection strategy during the life cycle.

In this annual report of the project “Experimentally verified model based predictions for integrity of copper overpack” (PRECO2017) the status of the experimental and modelling activities and results has been summarised up to the end of January 2018. The testing program and related activities have been continued and partly modified to support the assessment and verification of the long term integrity of the protective copper (Cu-OFP) overpack of the canister. To consider more discontinuous uniaxial behaviour, creep tests with stepwise increasing load and stress dips were conducted, and the results were compared with those from normal constant load creep testing. In general, the transients especially from stepwise increase in loading can consume a part of the deformation capacity and shorten creep (or creep-fatigue) life. The relaxation testing programme has continued with cyclic relaxation tests. The relaxation testing aims at developing a relaxation model for the FE calculation. It is expected that with the relaxation model the stress and strain distributions will be different when compared to the FE results with a traditional creep model only. The results of continuing multiaxial (notched bar) creep testing program suggest notch weakening, or life reduction by tensile multiaxiality. The 10.000h testing of a CT specimen from the modified welding (FSW) process with argon protective atmosphere is complete and the specimen is going to be investigated by metallography to study the behaviour of the oxide particles.

The share of energy produced from renewable resources is rapidly growing. Future conventional power plants will increasingly shift from base-load to more fluctuating load-follow operation, to balance the supply and demand in the grid. This will mean more frequent and faster load changes, decreasing minimum loads and more start-ups and shutdowns than assumed in design. The resulting challenges can be severe for effective operation aiming to avoid excessive cost from increased rate of damage to the plant equipment and from increased emissions at part-load and transient operation. At the same time, fuel feedstocks are increasingly diverse, containing e.g. different types of biomass and low rank coals, allowing the use of local resources and reduced dependence on imported fuels. The fuel changes may not be directly visible in operational parameters, but fouling and corrosion tend to increase material wastage and risk of tube leakage, operational problems and unplanned shutdowns. The future roles of thermal power plants will also frame the economic conditions of their survival. This paper aims to describe the challenges and possible solutions for successful operation of flexible power plants.

The performance of materials in critical high temperature components of power and process plants is one of the major technical reasons limiting both the component life and process efficiency. New structural materials can be attractive in principle, but it has been challenging to simultaneously achieve good mechanical strength, oxidation resistance and cost-effective composition in the candidate alloys. The optimal material solutions for a given component depend on multiple factors, but one typical limiting factor is the waterside oxidation resistance at highest operating temperatures and pressures that extend to the supercritical (SC) or even ultra-supercritical (USC) range. To facilitate development of improved alloys for the new demanding applications, the methods of materials modelling can be useful to explore the effects of new ranges of chemical composition and fabrication routes. In the current paper oxidation modelling based on the concept of Cr and Ni equivalents is presented and applied to long term operation of superheaters made of nominally similar 11%Cr steel. The model quantifies the impact of the alloying elements on steam side oxidation to facilitate prediction of the oxide growth according to the alloy composition. The model can be used to predict the effect of variation of particular alloying elements, to explore the effects of alloying modification on performance for alloy development, and for life prediction when life is limited by water/steam side oxidation. The model prediction shows generally good agreement with the observed oxide layer thickness and also with prediction from conventional parabolic expressions when available for the alloy.

The paper highlights the practical aspects of new EU project SmartResilience and its application in Finland. The basic idea of the project is that modern critical infrastructures are becoming increasingly "smarter" (e.g. cities). Making the infrastructures "smarter" usually means making them smarter in normal operation and use, but it has to be checked if these smart critical infrastructures (SCIs) will behave equally "smartly" and be "smartly resilient" also when exposed to extreme threats, such as extreme weather disasters or, e.g., terrorist attacks. Similarly, the question is, if making existing infrastructure "smarter" is achieved by making it more complex, would it also make it more vulnerable? Would this affect resilience of an SCI as its ability to anticipate, prepare for, adapt and withstand, respond to, and recover? These questions are tackled by the project, which plans answer these questions in several steps: by (#1) identifying existing indicators suitable for assessing resilience of SCIs; (#2) identifying new "smart" resilience indicators (RIs) - including those from Big Data; (#3) developing a new advanced resilience assessment methodology based on smart RIs ("resilience indicators cube", including the resilience matrix); (#4) developing the interactive "SCI Dashboard" tool; and (#5) applying the methodology/tools in eight case studies, integrated under one virtual, smart-city-like, European case study. One of these case studies will take place in Finland, dealing with modern systems of energy supply.

Recurring waterwall leaks have been observed in a coal-fired boiler where subsequent inspections have shown through-wall internal oxidation, mostly at butt welds of 16Mo3 boiler tubes. In several cases the oxide had retained a thick pit plug even after local penetration of the tube wall. The original initiation site is no longer visible as it was removed by oxidation, but away from the most severe damage the internal oxidation appears to have initiated immediately downstream after a protruding weld root. This location, estimated oxidation rate and oxide appearance suggest that the mechanism is related to locally intensive heating in addition to flow turbulence. In the damage history of the same boiler wall, the first comparable leaks were recorded at shop welds with strongly protruding root shape, presumably inducing particularly strongly turbulent flow, and even then only after long term service. The most recent failures, however, occurred at normal repair welds or even within parent material, and after much shorter time in service. The characteristic features and causative factors of the tube failures are discussed in the paper, based on the available evidence and experience from operation and maintenance.

Steam driven turbogenerator units with size up to nearly 2 GW provide large fraction of the electric power in the world. Steam turbines may show good reliability, but their disturbances tend to be associated with significant cost so that preventive and corrective maintenance is generally easily justified. However, it is often not simple or convenient to implement repairs or replacements when unexpected damage is found at the time of opening the casing. This is because many turbine components are large in size or unique by design, and possibly only available from a limited number of suppliers and even then with a long lead time. Fortunately alternative measures can be often taken, and two types of steam turbine damage are considered here: steam erosion of inner casings and other stationary components subjected to flowing wet steam, and steam leaks of the main flange surfaces. From the user point of view, it may matter that much of the manufacturers' knowledge on turbines resides far away, partly because unlike some decades ago, the number of manufacturers in the world is now small. Together with low number of comparable machines, this may to some extent limit the local availability of expertise and experience for successful maintenance. It helps however that many features such as materials and operating conditions in steam turbines are not particularly daunting in the required knowledge accumulation, and that today the manufacturers may agree to offer help in the turbine maintenance regardless of the original manufacturer. Challenges may appear when outsourced turnkey maintenance implies shipping of major turbine sections or components to factory, to correspondingly extend the shutdown period because of the geographic distance involved, even if the maintenance project were to stay within schedule. It is demonstrated in the paper that successful on-site repair is often a viable and attractive alternative to conventional full repairs or replacements.

Risk-based inspection (RBI) has become common approach to support asset management in e.g. refineries, chemical industry in general and a few other industrial sectors. Although there are regionally established traditions to use RBI also in power plants, recently the interest has been spreading to include a widening base of common power plants. This is partly due to increasingly established standard practices, tools and availability of the techniques also for smaller scale users, and partly due to the generally established benefits and needs for safety and optimization of resources. In this paper we show how RBI was applied in a small scale for the main steam line of a CHP (combined heat and power) plant in Finland. The experience showed some typical features of initiating RBI, for example how important and beneficial (although at first somewhat tedious) it is to systematically gather and handle the required background information of design, fabrication, inspections, maintenance and operational history for RBI of an ageing plant. In the case of the example plant, the steam system had been significantly modified because of process changes, and the level of documentation was not equally thorough everywhere. Existing conventional guidelines and experience were found to be useful to support relatively lean RBI by suggesting typical areas of elevated probability of failure and therefore recommended areas to be inspected, for example at welds of large branches, locations of high strains in thermal system displacement or deviations in the function of the piping supports. Observed indications in the inspections were largely consistent with expectations and only resulted in repairs within the planned outage period. The resulting risk-based (or -informed) inspection and action plans are the main outcomes of the RBI exercise, aiming to produce the best justification and return on the maintenance investment, and to support safe operation and long-term lifecycle performance of the plant.

Risk-based approaches can help in selecting and prioritizing future choices and schedules so that decisions are justified by the associated risk. The paper will give examples of applications from conventional power plants, where degrading physical condition can increase the risk of failure and decrease asset value: a single detailed example illustrated the approach, which is also anchored in the new European standard EN 16991 (prEN 16991:2016) and applied in large-scale industrial projects world-wide. To manage the risk, implemented inspections and maintenance can help to limit the probability of failure (PoF) and its uncertainty. Failure statistics of similar components operating in similar environments may provide fair estimates of PoF, but biased or lacking comparable statistics may require adjustment through e.g. additional measurements or expert information for the application. In parallel, inspection and monitoring data can more directly help in assessing the evolution of PoF. An example of the latter is a power plant heat exchanger, with inspection and failure criteria considered for a full life cycle that included multiple inspections by the same inspecting party and equipment. Uneven spatial distribution of in-service wall loss without changes in mechanisms or loss rates at a given location nevertheless required a change of inspection strategy during the life cycle.

This report describes candidate resilience issues and indicators to be used when assessing, predicting and monitoring resilience of Smart Critical Infrastructures (SCIs). A total of 233 candidate issues and 1264 indicators are provided for various threats, SCIs and the five phases of the resilience cycle used in the SmartResilience project. Structured candidate issues and indicators are mainly provided by collecting existing issues/indicators from the risk, safety, security, crisis management, business continuity and similar domains, considering resilience as an "umbrella". The main intended reader is the person (user) - within a city or area, or a specific SCI - who is responsible for performing the resilience assessment, prediction or monitoring, including carrying out necessary calculations. This can be an in-house person performing e.g. selfassessment, or it may be an external assessor. The issues and indicators presented in this report and stored in the SmartResilience database are candidate issues and indicators. Each user is responsible for finding a relevant and complete set of issues and indicators for his/her own case study. The report may be used together with the SmartResilience methodology and the Guideline for assessing, predicting and monitoring resilience of SCIs developed in the SmartResilience project.

The report discusses the challenges posed by four types of threats -terrorist attacks, cyber attacks, extreme weather and social unrest- on the SmartResilience case studies. The way this analysis was conducted was by assessing these threats using a 5x5 framework matrix. The two axes of the matrix were phases (understand risks, anticipate/prepare, absorb/withstand, respond/recover, adapt/learn) and dimensions (system/physical, information/data, organizational/business, societal/political, cognitive/decision-making). Each individual matrix block was discussed by subject experts who identified specific challenges and implications for each matrix element and rated its relevance (high, medium, low). In terms of the results, the system/physical dimension received the highest number of important challenges. Overall, the most important singular element was to understand risks in the organizational/business dimension. The least importance was attributed to the adapt/learn phase.

Tubular heat exchangers (HE) are common and often physically large in power and process industries, e.g. in steam generators, boilers, preheaters, furnaces and many other essential applications to transfer heat of process fluids. HE performance is important for managing the overall efficiency, economy, life and predictable maintainability of the process & plant: typical investment cycle ~ 20-60 years, opportunity for planned inspections and maintenance once every ~ 1-6 years. Intervening damage or failures will reduce equipment life and increase the risk of unplanned shutdown/maintenance, business interruptions and high cost recovery. Systematic countermeasures include risk-based methods, such as risk-based inspections (RBI).

Riskipohjainen tarkastus (RBI) tahtaa riskien mukaan priorisoituun tarkastusten kohdentamiseen ja ajoitukseen rakenteissa etenevan tai epaillyn vaurion (ikaantymisen) varalta. Asianmukainen tarkastus vahentaa arvioitua riskia pienentamalla vauriotodennakoisyytta tai ainakin sen epavarmuutta. Tassa tyossa on kasitelty esimerkkina kaytetynkaukolammonvaihtimen tarkastustietojen hyodyntamista RBl-nakokulmasta. Lammonvaihdinputkien tarkastus- ja vauriokriteereja on tata varten tarkasteltu yhden elinikasyklin eli uudelleenputkitusten valiselta ajalta, johon on sisaltynyt kuusi saman ta.rkastustoimittajan ultraaanitarkastusta. Lammonvaihtimen putket ohentuivat kaytossa epatasaisesti, mutta eivat hallitsemattomasti putkien ulkopinnan pisaraeroosion vaikutuksesta. Tarkastukset keskittyivat lyhimman elinian kokeviin alle tuhanteen putkeen, joista kasvava osa tulpattiin elinikasyklin kuluessa vuotojen ja alasajojen valttamiseksi. Putki- ja aluekohtaisesta ohenemisnopeudesta huolimatta maksimiohenemien jakaumat olivat yhdenmukaisia ja oheneminen eteni ennustettavasti samalla mekanismilla ja odotetuissa mittausvirherajoissa. Arvion mukaan kaytetyt tarkastus- ja toimenpidekriteerit ovat olleet asianmukaisia seka ennakoivan kunnossapidon eli oheneman ja tulppausten kannalta etta vaurioriskin eli kayttokeskeytysten(vuotojen) nakokulmasta. Onnistumisia kuvaa se, etta ensimmaista vuotoa seuranneella toimenpideohjelmalla uusia vuotoja ei esiintynyt seitsemaan vuoteen.

The ferritic-martensitic steel P91 (X10CrMoVNb9-1) is widely used in high temperature steam systems of power plants, and it is a candidate material for Gen-IV reactors. In comparison to austenitic steels, thick-wall P91 has relatively attractive mechanical and physical properties combined with resistance to stress corrosion cracking in water-steam environments. This study aimed to explore the combined cyclic, creep and relaxation behaviour of P91 material up to a component wall thickness of 60 mm. Uniaxial specimens were subjected to cyclic loadings with periodic forward creep or relaxation at peak stress. The results indicate that prior creep or intermediate relaxation periods up to 72 h will influence the subsequent cyclic softening of P91, but do not significantly reduce the cyclic life. In contrast, prior cycling has a detrimental effect on the subsequent creep life. A simplified creep-fatigue model is shown to predict life better than usual code-based approaches for cyclic service of P91 steel. Improved verification of all models would benefit from the availability of more extensive long-term data on P91 steel.

For higher efficiency and competitive overall performance, it would be an advantage to be able to safely exceed the highest operational values, currently up to about 600-620°C/25-30 MPa in supercritical (SC) boiler plants. Under such operating conditions the oxidation resistance in SC water tends to limit the potential for further improvement of efficiency. The way to increase the oxidation resistance of traditional austenitic boiler tubes e.g. TP 347H is to do additional cold work on the boiler tube inner surface. In the current paper the effect of cold work on the oxidation resistance of TP347H and TP347HFG has been evaluated by shot peened samples with different parameters and subjecting those samples to supercritical oxidation exposure. The results show an improvement in the oxidation resistance of the alloys, especially in the large grained alloy TP347H. Also the uniformity of the deformation layer was seen to have an influence on the oxidation resistance, since the oxide nodules start to grow at locations with the thinnest or no deformation layer.

Steam turbine is a rotating machine and therefore fatigue is an important damage mechanism to consider already in turbine design, and it is remains necessary to consider the risk of initiation and growth of fatigue cracks in the later in-service inspection and maintenance activities. With successful design and fabrication, and operation within the intended limits, the expected fatigue life is relatively long but nevertheless finite and variable. The design features and materials of steam turbine blades are generally well established, and conventional methods of inspection and maintenance are widely applied to monitor and limit the evolution of in-service damage. As the fatigue performance of a blade is simultaneously sensitive to material characteristics, structural/loading geometry and surface condition, the combined potential range of these factors explains in principle the observed wide scatter in turbine blade life. In this paper, examples on reduced blade life are discussed based on recent experience with a case example.

Risk-based inspection (RBI) can offer justified priorities in timing and targeting of in-service inspections and maintenance action for power and process plants. For this purpose, recommended procedures and tools are increasingly supported by the evolving standardization on risk based assessments and RBI. Successful RBI generally requires availability of background data on design, fabrication, operation and maintenance history, evaluation of the current condition and its expected deterioration in systems and components of interest, and expertise of the RBI team to conduct the assessments. This paper aims to show some RBI related observations, modelling and experiences when considering the accumulating damage, life-limiting mechanisms and failures in high temperature components of thermal power plants. An example is shown from a pilot scale application of RBI on a steam system of a combined heat and power (CHP) plant after more than 200 000 h of service. The implications and experiences are discussed in terms of the implemented inspection and maintenance actions, and perceived benefits of RBI.

A description of creep strain evolution is frequently needed in design or life assessment of components subjected to service at high temperatures. Unfortunately, long term creep strain data are not as easily available as rupture data or rupture models. One of the most demanding tasks in this context is to predict the creep behaviour reasonably accurately beyond the range of available data. Much effort has been invested into developing reliable methods to extrapolate creep rupture data, for example in the recommended procedures of the European Creep Collaborative Committee (ECCC) and PD6605 of BSI. However, for strain no such tools are currently available. Here a new and robust creep strain model is suggested to provide the whole creep curves based on the corresponding creep rupture model. This logistic creep strain prediction (LCSP) model defines the creep curve only with three additional parameters to those of the corresponding rupture model. In its basic form the LCSP model optimises a non-linear asymmetric logistic transition function fitted in logarithmic strain against a time temperature parameter (TTP), giving time to specified strain. Unique features of the model include its simple inverted expression for strain and strain rate. The fitting effectiveness of the new method is shown to match all the contesting creep strain models of the ECCC intercomparison activity on a single heat data set of the steel P22 (10CrMo9-10). The model is also shown to produce accurate predictions of the stress to 1% strain up to 100 000 h, when compared to the values given in DIN 17243.

Power plant structures operating at high temperatures are designed for a lengthy explicit or implied minimum life, dictated by the expected life-consuming damage mechanisms like corrosion, creep, fatigue and other forms of wear and tear. The actually realised damage shows highest rates at locations of the most adverse combination of material properties (weakness e.g. in welds), loads (mechanical, environmental, thermal) and other features like geometry and deviations from the intended condition. In case of no additional surprises, the technically justified life is likely to nevertheless differ from the assumptions in design. The explicit or implied "safety factors" will offer on average some extended life and service potential to components even after operation up to and beyond the nominal design life. To avoid surprises, it remains wise be well informed of the evolving material and component condition of an ageing plant. The common questions are when and how should one measure to achieve the required confidence for decisions to run, repair or replace. In this paper we contemplate the challenges for managing the life of critical components of current and future power plants. Radical changes may happen in the merit order of plants by e.g. shifting fuel prices or public policies, and this could naturally affect the need, available budget and planning to maintain a given plant. Less impact is expected in the established principles of the technical operations like monitoring, inspections and interpretation of the results. However, changes are likely in selected areas related to materials and component lifing.

Storing solid fuels like coal in large quantity is known to involve a risk of spontaneous combustion. In comparison to conventional open stockpiles, closed storage may in principle decrease this risk by more limited ingress of air, but the risk can also be enhanced for example by reduced access for preventive or mitigating action. The risk of self-heating until ignition is considered here for a unique underground rock storage of a coal-fired power plant. The main adverse consequence of non-functioning storage is interrupted district heating that is more difficult to replace in winter than lost electricity supply. To predict the risk of spontaneous combustion with reasonable confidence, material modelling of coal properties has been applied with verification by comparison to actual field evidence. Fire risk management is discussed in terms of the available early warning indicators and options for cooling, extinguishing and prioritised use of stored fuel.

At high temperatures the water/steam side oxidation resistance is an important issue for boiler components like superheaters, where the growing internal oxide decreases heat transfer and increases surface temperature. This is increasingly critical for modern plants that aim for high efficiency from elevated levels of operating temperature and pressure. In this paper, a new simple model is presented to describe the steam/water side oxidation rate of common alloys for the high temperature end of the boiler plant. The model is based on the chemical composition of the alloys, applying equivalence expressions for collapsing the multi-variable problem to a more easily manageable setting with a combined parametric oxidation (POX) model. For verification, a range of experimental water/steam oxidation data from numerous tested alloys has been compared with the model predicted oxide growth. The results suggest a surprisingly good agreement with the model and test data. Although the optimised model parameters differ between e.g. steels and nickel alloys, as expected due to difference in the oxides, they remain very similar within a given material class, e.g. for all tested steels.

The advantage of using supercritical water (SCW) systems for power generation is based on the increased thermodynamic efficiency when operating at higher temperature and pressure. This has been realised for many years in medium to large size supercritical fossil plants. High efficiency in power generation is not only desirable because of economical reasons but also for enhanced environmental performance meaning reduced quantity of forming ash and emitted pollutants including CO2. Steam oxidation has become an important issue for steam power plants as operating temperatures increase from current to 650°C and even higher. To achieve these higher steam values new materials are needed. This paper presents results of the oxidation performance of potential new alloys (FeCrAlY, NiCrAl and Sanicro 25) in a supercritical water autoclave environment at 650°C/250 bar, with 150 ppb dissolved oxygen.

The research programme on the creep issues of the nuclear waste disposal canister copper material has continued with long-running uniaxial and multiaxial experiments as well as relaxation testing to support the material modelling activities, which produce appropriate material models to be used in the FE analysis, which will be used to assess the stresses and strains in the critical locations of the canister. The relaxation testing is ongoing and it has proven necessary to use a servo-mechanical testing machine in order to generate good quality test data. The initial FE prediction of strain rates calculated by the LCSP creep model and a relaxation model suggests a clear difference, but this needs to be verified after a better relaxation model is defined based on much more data. The most important result of this year has been the observation of cracking in the oxide particle zone at the friction stir weld root in the longest running CT test, which was terminated after 50 000 hours for sectioning and metallographic investigations. The oxide particles have nucleated creep cavities which have eventually lead to cracking of the oxide particle zone in spite of the low stress state 8 mm ahead of the notch tip. This observation is in line with the earlier results from uniaxial creep tests with a radial sample from the weld root. The weakness only manifests itself in long-term testing because the cavity nucleation and growth process is a slow process. Because of the low strength of the oxide particle zone welding in air should be avoided and vacuum or inert gas should be used instead.

Risk-based inspection (RBI) is not always applied in power plants that are generally not seen to carry similar safety and other hazards as e.g. some branches of the chemical industry. Hence the supporting standardization, industrial guidelines and available services for RBI activities have not been as well established in the power industry. Such support has been addressed for example in the European projects RIMAP, iNTeg-Risk and SafeLife-X, with results that include recommendations towards the evolving European standardization on risk-based assessments. This paper aims to explore the benefits of probabilistic life assessment in supporting RBI and related maintenance planning. The details in the input information, including the progression rates of the active damage mechanisms, will affect the associated uncertainty in the probability of failure, predicted minimum life, and the resulting conclusions on next inspections. However, successful probabilistic treatment of the input information can allow realistic treatments of the models for the indicated or postulated damage mechanisms, and by using for example Monte Carlo approach, the outcome prediction is not sensitive to the mathematical formulation of the individual or combined damage mechanism models. The examples have been selected from components subjected to high temperature service in applications where one can expect unavoidable damage accumulation in long term operation, and/or multiple mechanisms affecting the component life. Implications are discussed in terms of implementation in RBI and other aspects of managing the assets.

To characterize the actual local long term operating conditions, possible deviations in them and the material condition, it is essential to be able to understand and evaluate the time and temperature dependent change of high temperature structural materials of power and process plants. In this paper the thermal degradation of selected steels is discussed with comparison of materials extracted from plant after long term service and similar materials thermally exposed to simulating heat treatment (controlled annealing). In general, such comparisons can be used to assess for example the effective in-service temperatures. However, there are also caveats to consider, such as variation of the initial composition and heat treatments, impact of stress-activated component of diffusion, and the material-dependent differences in the kinetics of different aspects of microstructural degradation. Examples are given to illustrate the observed influence of these factors of uncertainty.