Assessing residual life of high temperature components


High temperature pressure parts of steels which have aged beyond the designed life after long service in high temperature and high pressure steam service in  boilers and petrochemical steam process plants need remaining life assessment. The high temperature steel components undergo irreversible creep damage. Creep is a steady and irreversible material damage of metals at high temperature and stress. For high temperature alloys as in main steam pipelines working at 450 - 600oC, creep governs the major damage mechanism.

The material degradation due to creep will have to be evaluated on each component at several critical locations. The theoretical approach of calculating remaining life is based on temperature, stress and time of operation using numerous computer software has been found highly erroneous. It has been found that estimation of remaining life by destructive testing by taking part of the component and performing accelerated uniaxial creep test will lead to several uncertainties and errors. In such laboratory tests, it will not be possible to simulate low cycle creep fatigue, system bending stresses and biaxial stress. Besides, cutting part of a component and welding it with a matching material will lead to additional creep strain and damage, resulting in further reduction of the existing life of aged component.

Destructive tests are seriously limited to a apart of a component, and the testing duration is high and the cost involved in the entire work is excessively large as compared to in- situ metallography and replication. The latter evaluation can be done on numerous components in a short time, and the reliability has been acclaimed as equal to accelerated creep tests of low alloy steels. However, creep damage involving detection of fine creep pores (CP) and their estimation require high quality in- situ metallography.
In a practical evaluation of various techniques, it has been found that in-situ metallography and replication are highly sensitive in detecting various levels of creep damage in high temperature ferritic alloy steels. 

Two to three replicas are taken for each spot: to be analysed using scientific methods, gold sputtered and studied using optical and or a electron microscope.

If no creep pores are seen, but only spheroidized carbides are observed, then creep pores is considered  less and the next inspection is needed after six years. According if isolated creep pores are seen further inspection is needed after three years and a decision will be taken thereafter regarding further service. 

It has been calculated "Neubauer" that even if a few creep pores are seen in replica at 500x, 50 per cent of the life is consumed by the service conditions in low alloy ferritic steel steam retaining pressure parts in thermal power plants. 

Another quantitative approach is based on the fraction of cavitated grain boundaries "A'' in the typical observation of replicas. It is relatively easy to count the total number of grain boundaries in a field of observation and also the spoiled grain boundary by a creep pore. It is somewhat conservative and semi quantitative in approach.

Creep damage in steam pipes can cause two types of failures. The first is ``leak before rupture'' and this occurs where there is high deformation potential as in straight portions without welds. The second is ``rupture without leak''. The heat affected zone regions of long seam welds, circumferential weld, pipe bends and Tee section welds fail without significant leak after the final stage of material exhaustion. In fact, the consumption of total available deformation potential reduces remaining life.                   In regions with low deformation potential, diffusion creep predominates over dislocations whereas in the areas of high deformation potential, dislocation controlled creep is more predominant.

At high temperatures in the creep range, the material deforms also by grain boundary sliding under the application of stress. Such phenomenon occurs by complex dislocation involving climb and annihilation at grain boundaries. The crystalline engineering materials (eg. alloys of Fe, Al, Ti, Ni) at high temperatures under constant stress gradually expand in three stages ending in rupture of the material. They are primary creep, secondary or steady state creep and secondary or steady creep and tertiary or accelerated creep. The microstructure changes drastically after creep service. 

The percentage of elongation after creep is very small near the heat affected zone (HAZ) of the weld, for example, circumferential and longitudinal welds in steam pipes of alloy steels.

Creep damage is accumulated rapidly in HAZ due to fine grain size and complex factors involving carbide precipitation and welding residual stresses. The in-situ microstructural study of the creep-serviced component at site can reveal the extent of creep damage and can also indicate the remaining (useful) creep ductility, which is a major factor to decide the remaining safe life.

The safe remaining life time or RL of high temperature welded steel components in fossil power plants which exceeded the design life for example, 100,000 hours (11.4 years) or 200,000 hours (22.8 years), is decided not by retirement-by-time basis, but by retirement-for-cause strategies, that is the actual material degradation due to creep will have to be evaluated on each component at several critical locations such as main steam pipe, exit superheater header with several welds and welded fittings.  The designed life is based on the lower bound data of creep curve, with additional factors of safety index.

Hence, depending upon the quality of steel making practices, fabrication processes and plant operating conditions, remaining life is usually available in greatly varying degree. Destructive tests are limited to a part of a component, and the testing duration is high and the cost involved in the entire work is excessively large as compared to in-situ metallography and replication.

The latter evaluation  can be done on numerous components in a short time and the reliability has been acclaimed as equal to accelerated creep tests of low alloy steels. However, since creep damage involving detection of fine creep pores  and their estimation require high quality in-situ metallography in the sense that plastic deformation due to mechanical polishing can cover up creep pores leading to an underestimation of creep damage and such errors is avoided by metserve advanced sample preparation.

The high temperature ductility of a metal under constant load leads to deformation by creep mechanism. Three stages of creep deformation occur leading to distinct microstructural change in five stages.  This is a topic of global importance for the life assessment of serviced high temperature components in power and process plants.

 

The computer modeling and calculation of Residual Life has serious errors as high as 400 per cent due to uncertainties in stress state, microstructure, creep-fatigue conditions etc. and hence are considered unfavorable.
wpe5.jpg (35981 bytes)Thus, of the various methods, metallographic method has come to stay and is the one favored widely all over the world in view of its several advantages over the other methods. Apart from direct examination of the microstructural condition of the plant component in-situ, the technology of replication of the micro- structure with very high fidelity for examination in laboratory has been developed and widely practiced. The plastic replication technique offers several advantages like high quality equal to or better than that of direct metal examination. Being non- destructive, it is useful for periodic monitoring of plants at specific intervals. The replicas serve as a permanent record of observations and can be stored. Replicas can be examined at high magnifications in SEM with a very high degree of resolution.

The detailed RLA report is reviewed by statutory authorities ,this approach helps obtain full value for the money spent on the component by actual assessment. From experience it has been found that sufficient safe remaining life existed for numerous components, and this will help to run the plants by utilising the full remaining life.