Problems of ageing equipment residual life assessment
Dr., Professor A.A. Dubov
The problem of providing the reliable operation of equipment, vessels, gas and oil pipelines and various structures becomes more and more relevant every year as the equipment ageing in many branches of industry significantly surpasses the rates of technical re-equipment. For instance, in power engineering as of September, 2002, about 90% of thermoelectric power stations equipment had exhausted its park life and the significant part of it had achieved physical wear. The above-mentioned problem is aggravated by the lack of scientifically grounded concept of technical diagnostics and life determination and by insufficient effectiveness of conventional methods and means of metal non-destructive testing.
Based on the analysis of existing approaches to the ageing equipment residual life assessment formed in various branches of industry, the following general trends can be marked out.
Firstly, many specialists in the sphere of equipment reliability pass from probabilistic methods of life assessment based on failure statistics to assessment of individual life of the ageing equipment base on the complex approach combining the results of destructive and non-destructive testing with calibrating calculations of strength.
Secondly, at life assessment a tendency has been noticed of shifting from crack detection to technical diagnostics methods based on combination of fracture mechanics, physical metallurgy and NDT. Equipment and structures stress-strained state NDT methods come to the forefront.
Thirdly, the necessity for a 100% examination of the ageing equipment aimed at determination of potentially dangerous zones has been realized.
At the same time the following drawbacks and defects existing at realization of these approaches should be noted.
At a complex application of various methods and means of non-destructive and destructive testing there is no strictly specified order and sequence of their application for a specific test object.
As it is known, the order, the scope and the frequency of equipment inspection is determined, on the one hand, by the park (design) life, damageability, overhaul life and, on the other hand, by availability of inspection methods and means and their capabilities.
Special instructions on the order and frequency of inspection and prolongation of equipment life are available only in certain most critical branches of industry (for example, nuclear-power engineering and heat-power engineering) [1, 2, 3]. And even in these advanced fields (from the viewpoint of arrangement of the equipment metal state control) there is a problem of metal limiting state determination and the equipment individual life assessment .
The suggested methods of strength calibration calculation can be conditionally divided in four groups:
methods of calculation by metal corrosion rate;
metal crack resistance calculation methods;
metal fatigue calculation methods;
calculation methods for equipment units operating in conditions of creep.
The main defect of well-known methods here is that they suggest a low level of permissible stresses [σ ]. As a rule, the level [σ ]?σ 0,2/2, where σ 0,2 - is the conditional metal yield strength. There is a requirement in level calculation for critical structures [σ ]<0,3σ 0,2. It is known that these requirements are governed by the equipment metal work in conditions of glide and by shear strain. As the practice shows, these conditions of metal work are determining for the structure reliability. However, it is impossible to predict in advance the zone of metal glide areas on the equipment using calculation methods.
Besides, the existing strength calculation methods assume, as a rule, independent flow of corrosion, fatigue and creep processes, though in practice these processes flow simultaneously in various combinations.
The tendency of shifting from traditional crack detection to technical diagnostics using the complex approach incorporating: defect parameters determination, internal (residual) stresses distribution assessment, determination of actual structural-mechanical characteristics of metal, is restrained, first of all, by the low effectiveness of the current methods and means of equipment stress-strained state control. For example, the paper  states that at a current stage none of the tested means of stress determination (about 10 various stress inspection instruments were tested) can provide authentic data on the stress-strained state (SSS) of gas pipelines in real operational conditions.
The analysis of the known inspection means capabilities and stress measurements in the base metal of equipment and structures weldments and welded joints allow naming their major drawbacks. The basic drawbacks are:
impossibility to use most of methods in the plastic strain area;
control locality, their unsuitability for long structures inspection;
metal structure change is not considered;
inspection is carried out only on the surface of weldments, impossibility to assess the depth layers of metal and welded joints metal;
the need to make graduated diagrams based on preliminarily prepared samples;
the need for test surface and test objects preparation (dressing, active magnetization, sensors adhesion, etc.);
complexity of testing sensors location determination related to the direction of the action of main stresses and strains determining the structure reliability.
It was noted earlier that stress concentration zones were the main sources of damages development. The metal structural-mechanical properties need to be first of all investigated exactly in SCZ. The existing traditional methods of stresses non-destructive testing (X-ray, ultrasonic inspection, Barkhausen noise and others) do not allow solving this complex problem of SCZ determination on equipment due to operational loads action.
Though the necessity for 100% equipment examination at life assessment is realized, however, much time and large material and financial costs are required for implementation of this task in practice. This task is not realized in practice using conventional NDT methods (ultrasonic inspection, X-ray, magnetic particle inspection). For instance, the length of pipe heating surfaces on a modern 1000 t/h steam boiler makes more than 500km. Therefore it is practically impossible to tap, clean and measure by ultrasonic inspection method such a number of pipes, and none of electric power stations does this work. Similar problems occur at inspection of gas and oil pipelines the length of which in Russia reaches hundreds of thousands of kilometers, in petroleum and chemical industries at inspection of a large park of vessels and pipelines as well as in other branches of industry at inspection of ageing equipment and structures.
Let us consider further the capabilities of current (conventional) NDT methods and means at solution of tasks occurring at equipment life assessment.
The existing conventional NDT methods and means (ultrasonic inspection, magnetic particle inspection, X-ray) are known to aim at searching and detection of a specific defect. Determination of the size of defects (occurrence depth, length), located in the volume of the base metal or in the welded joint metal is a complex practical task. However, if the size of the defect is determined (modern crack detectors solve this task), it is necessary to determine the extent of this defect danger and to answer the question: "Is this defect developing or not?". To answer this question a calibration calculation of this unit strength should be made taking into account the defect size. It is obvious that such calculations are not carried out in general practice. Therefore the existing norms on defects permissibility (revealed by ultrasonic inspection, X-ray) for instance, in welded joints are mainly based on statistics and in most instructions have conditional nature. There are no scientifically grounded norms on defect size permissibility from the viewpoint of fracture mechanics in the general practice.
If capabilities of, for instance, magnetic particle inspection and eddy-current control methods, aimed at surface cracks detection, are considered, the following should be noted here. Despite the fact that the modern instrumentation and testing technology using the indicated methods has been significantly developed nowadays, there are till date no norms on surface defects size permissibility for equipment in operation in many branches of industry.
The existing norms and samples used, for example, in magnetic particle inspection, were developed for new machine-building products. These norms are nor suitable for equipment in operation for the following reasons: firstly, the slag, the metal external layer corrosion do not allow applying the indicated control means and methods without cleaning and removal of this layer, and secondly, these norms from the viewpoint of fracture mechanics require special grounding practically for every test object. Therefore for the critical equipment in operation, for example, at thermoelectric power stations surface cracks on most test units are not allowed and should be removed . Thus, samples and norms specified in instructions for magnetic particle inspection and eddy-current control methods are applied in general practice as a measure of sensitivity of the instruments used.
The tasks of internal defects control in fillet, branch and T-joints, in contact welded joints, in small-size joints (up to 6 mm), determination of corrosion pits on pipeline internal surfaces are complex and not till date solved by traditional crack detection methods.
Unsuitability of conventional NDT methods for defects detection at an early stage of their development should be noted as well. More and more specialists start realizing that "pre-defec" metal state is in many cases (especially on the ageing equipment) more dangerous, when irreversible changes took place at a structural level and the fatigue-assisted damage may occur all of a sudden and, as a rule, in unexpected zones. The sensitivity level of conventional NDT methods does not allow revealing the "pre-defect" state of a metal.
Methods and means of metal structural-mechanical properties NDT (measurement of hardness, coercive force and of other magnetic characteristics of metal, "replicas" taking for structural changes determination and other methods) are widely used at equipment life assessment nowadays. Complex methods of metal’s physical-mechanic properties NDT are developed and being applied in practice, for example, plants for combined application of magnetographic method and kinetic indenting method , the Moscow Power Institute’s instruments and methods for materials testing by indentation or scratching for rapid evaluation of mechanical properties  and others.
At present there are about 20 standards for non-destructive and partially destructive sampling methods in Russia. All the available standards determine the sampling mechanism, i.e. answer the question: "How to carry out sampling?". This variety does not contain a single standard answering the question: "Where to take a metal sample from?". Therefore at carrying our sampling on the equipment after long operation to assess metal degradation specialists make conclusion on the metal state only at the place of sampling. It is impossible to extend the results of this conclusion on the entire metal of the test object (and even of an individual element, for instance, the steam-water pipe bend). Metal samples are taken, as a rule, from zones of the most probable development of damages (or from zones where metal damages already existed).
It was noted earlier that SCZ, occurring at the stable dislocation slipbands zones and caused by the action of operational loads, are the major sources of equipment damages. According to the inspection experience, these zones on the equipment metal surface show themselves in the form of lines with the size by width and depth at the beginning of their development of not more than several microns. The probability to hit these zones at metal sampling is very low. It is obvious that such task can be solved only at a 100% metal examination on the entire surface of the test object using highly sensitive methods. There were no such methods enabling to solve this task till date.
In this connection it should be noted that if there is no opportunity to determine SCZ and to carry out metal sampling, then, accordingly, the intention to make strength calibration calculation for residual life assessment loses its sense. Only in exclusive cases, when, for instance, the metal is affected by corrosion with pipe wall (or vessel shell) thinning on a large area, it makes sense to calculated strength taking into account wall thickness and corrosion rate decrease.
Thus, the presented brief analysis of existing methods of metal damages and degradation NDT demonstrates their low effectiveness at industrial equipment life assessment. The tendency of shifting from traditional crack detection to technical diagnostics using principally different control methods and approaches becomes clear and appropriate. More complicated tasks occurring at equipment life assessment (as compared to conventional crack detection at normal operation) require application of means and methods that are more difficult to master but more effective at control of altering metal properties. First of all, means and methods allowing practical control of equipment’s stress-strained state should be assigned to such methods.
All leading diagnostic centers of the world are occupied nowadays by the problem of mechanical stresses measurement in operating structures in order to assess their state. However, no effective methods of stress control, suitable for practical application, have been suggested till date.
Major drawbacks of traditional stresses and strains NDT methods were marked out above.
Besides, traditional methods and means of stresses NDT based on active interaction of the instrument signal with the structure metal obtain indirect information on the test object’s stressed state, i.e. have insufficient self-descriptiveness of physical fields used at control.
Indeed, the introduced into the investigated material field, interacting with the material’s proper fields, alters its properties and the test object’s stress-strained state characteristics. The alterations life nature, amount and time are determined by the dynamic relationship of interacting fields’ energies. In practice at carrying out diagnostics such alterations are simply neglected.
Thus, the above listed drawbacks of the known SSS control methods are caused not only by metrological peculiarities, but also to a certain extent by these methods’ physics, i.e. they are regular. Lack of metrological basis for materials’ SSS characteristics measurement certification and calibration (there are no unified standards and samples in Russia and abroad) leads to requirements ambiguity and wrong methodological approach to the developed control means.
Paper  states that thermodynamic constitutive equation of solids must be taken as a basis of equipment reliability theory and prediction. Basic physical effects accompanying the metal fracture mechanism: mechanic, thermal, ultrasonic, magnetic, electric and electromagnetic are determined. It implies that using one or simultaneously several control parameters, reflecting the listed effects, it is possible to most objectively assess the test object’s stress-strained state.
It was stated earlier that the equipment metal work is mainly determined by dislocations glide and shear strain. The metal fatigue damaging accumulation in many cases occur in conditions of low- and multicycle operational load. The question is, how the traditional methods of stress control can asses actual structure SSS, when in the general case stress concentration zones due to shear strain are unknown. It is obvious that only "passive" methods of SSS diagnostics are able to answer the questions put and are the most suitable for practical application.
Passive NDT methods using radiant energy of structures, first of all, are:
acoustic emission method (AE);
metal magnetic memory method (MMM).
These two methods are nowadays widely spread in practice for early diagnostics of equipment and structures damages.
As it was demonstrated in practice, MMM, as compared to the AE method, gives additionally the information on the test object’s actual stress-strained state, which allows more objective determining of the reason for stress concentration zones formation, being the source of damaging development. Besides, application of MMM enables a 100% equipment examination with SCZ and defects detecting at an early stage of their development. Having the complete information on the revealed defects and on the possible influence of each of them on the equipment residual life, one can easily solve the task of recovery work scope determination necessary for improvement of units’ efficiency life to the required level.
Paper  gives the method for metal limiting state and equipment life determination using metal magnetic memory parameters.
1. GD 10-577-03. Standard instruction for metal control and lifetime prolongation of boilers, turbines and pipelines main units at thermal power stations. Moscow: ORGRES, 2003.
2. GD EO 0186-00. Method for Atomic Power Plant power unit vessels’ technical state and residual life assessment. Moscow: Concern "Rosenergoatom", 1999, 75p.
3. GD EO 0185-00. Method for Atomic Power Plant power unit pipelines’ technical state and residual life assessment. Moscow: Concern "Rosenergoatom", 1999, 63p.
4. The concept of power objects technical reequipment at RAO "Russian Unified Electric Power Systems" during the period till 2015. Document of RAO "Russian Unified Electric Power Systems". Moscow, November, 2001.
5. Dubov А.А., Demin Е.А., Milyaev А.I., Steklov О.I. Gas pipelines stress-strained state control // Gas industry, 2002, No.2, pp.58-61.
6. Matunin V.М. Methods and means of rapid assessment of structural materials’ mechanical properties without sampling. Moscow: Publishing House of Moscow Power Institute, 2001.
7. Komarovsky А.А. Stress-strained state diagnostics // Testing. Diagnostics, 2000, No.2, pp.22-27.
8. Dubov А.А. Means of metal limiting state and equipment life determination using metal magnetic memory parameters. Proceedings of the XVIth Russian Scientific-Technical Conference "Non-Destructive Testing and Diagnostics". St.-Petersburg, September, 2002.