The Procedure for Determining the Time of Safe Service beyond the Design Service Time Based on Creep Testing

This article describes the method for determination of residual life and disposable residual life based on long- and short-term creep tests for different temperature levels above the operating one and stress level equal to the operating one. The method for determination of the share of disposable life in total life and the time of safe service beyond the design service time was proposed. These characteristics of the steel are used for the evaluation of the structural changes and mechanical properties of the material after long-term operation. The result of this study is the database of material characteristics representing the mechanical properties related to the structure analysis, and it can be used for diagnosis of the components of pressure part of power boilers.


Introduction
The extension in the service time of power units which have exceeded the design service life of 100,000 h is made using the calculation method based on the creep strength data for 200,000 h service and the positive results of comprehensive diagnostic tests, with particular consideration given to those of the critical pressure components of boilers and turbines. Out of them, the components working at above the limit temperature, that is, under creep conditions, are of particular importance. In the assessment of these components, it is important and necessary to assess the condition of their materials [1][2][3][4][5][6][7]. Such an assessment is carried out based on nondestructive materials testing, and its result is referred to the available characteristics of materials after service [8][9][10][11][12][13][14][15]. The results of these tests provide good estimation of the condition of material and its exhaustion degree and allow for determination of the time of further safe service until the next inspection [16][17][18][19][20][21][22][23][24][25][26][27][28]. However, to reach the level of 200,000 h service for the material of components, not only good estimation of residual life but also its determination based on the destructive testing on a representative test specimen is required in a number of cases.
Yet, it is not always possible in practice. It can be made to assess the condition of the material of the main steam pipeline or transfer pipeline as well as for certain chambers and desuperheaters. However, it must be preceded by a cost-effectiveness calculation for such a procedure.
The problem related to the assessment of creep resistance in case of materials operated for a long time under creep conditions is the time required to carry out the creep tests to assess the residual life and the remaining safe time of service. This paper shows how to use long-term creep tests at the temperature and stress parameters corresponding to the actual operating ones, which allow for determination of the actual residual life, to make such an assessment. The method for using shortterm creep tests for such an assessment where test duration is reduced by increasing one of the parameters in relation to the actual ones is also presented.

Creep tests 2.1 Long-term creep tests
In the assessment of residual life and disposable residual life, the main source of information is still creep tests in spite of the fact that their disadvantage is a long time until the test results are obtained. Creep tests are usually carried out under uniaxial tension and at strain rates from 10 to 9 to 10-12 s À1 (corresponding to the actual service conditions) on test specimens taken directly from the material of component after long-term service at a temperature exceeding the limit one tg (te > tg).
Long-term creep tests can be divided into: • Creep rupture tests without measurement of elongation during test • Creep test with measurement of elongation during test Based on the results of creep rupture tests without measurement of elongation during test, the residual life can be determined using one of the two procedures: • Lifetime fracture rule

• Extrapolation of creep rupture test results
The lifetime fracture rule relies on the assumption that changes in temperature T or stress σ during service can be divided into degrees characterised by specific temperature Ti and stress σ i .
Thus, the lifetime fracture rule concept by Robinson assumes that where T is the time of stress σ at a temperature T, tr is the time to rupture under stress σ and at temperature T, and i is the number of degrees of stress and temperature T from the beginning of service until rupture of the material.
For each of these degrees, the creep rate in steady state (secondary creep) or the time to rupture tri based on the experimental data measured under preset stress and temperature is determined. The total creep strain is assumed to be the sum of strains corresponding to the consecutive degrees, that is, For the materials after service, the lifetime fracture rule can be presented as where t s is the time of previous service of the material, t rs is the time to rupture under service conditions (σ e , Te), t t is the time to rupture of the test specimen determined under the accelerated creep test conditions for the material after service (σ t > σ e and/or Tt > Te), and trt is the time to rupture of the material in the asreceived condition under the accelerated test conditions (σ e = σ t , T = Tt).
Assuming that the further operating conditions will not differ from the previous ones, the time equal to the residual life can be determined from The need to know the time to rupture of the material in the as-received condition t rt restricts the use of this method.
In the event when the test specimens for destructive testing are taken from the operated material after two different times of service ts1 and ts2 and the creep tests are carried out for both conditions of one material, the time to rupture of the material in the as-received condition trt can be eliminated. Then, the knowledge of the properties of material in the as-received condition is not required.
The use of this method is particularly useful in the modernisation of boilers when some critical boiler components operating under creep conditions are replaced and the basic operating parameters are recorded and controlled on-line.
The residual life can also be determined by extrapolating the results of creep rupture tests carried out on material after service at a temperature equal to or different from the design service temperature (Tr) and stress values equal to, but usually higher than, those characteristic of the service conditions (σ r ). Obtained in this way, the results of creep rupture tests with a time of usually approx. 10,000 h are presented in the parametric way by expressing one variable as a product of the function of two other variables, that is, stress and temperature and time, respectively, using the Larson-Miller parameter in the following form: where H (σ) is the stress function, σ is the stress, T is the temperature in K, C is the material constant at 1/T = 0, and t re is the time to rupture of material after service.
This method for determination of residual life is shown graphically in Figure 1. Knowing the operating temperature T r and the operating stress σ r (resulting from the component's geometry and operating pressure p r ) of further service, the time to rupture t re can be determined following which the component destruction should be expected. However, the determined time t re is of no practical value. The practical value is the time to end of secondary creep, which is called disposable residual life t re 0.6 and which is part of the residual life. This value is specific to each of the test grades of material.
The duration of the tests allowing for determination of the residual life t re can be reduced by carrying out the creep tests with measurement of elongation during test in order to determine the creep rate in steady state for the properly selected creep parameters (σb, Tb). Using the obtained curve of maximum creep rate vs. stress = f(σ) at the constant temperature Tep of further service, the time of further safe service can be determined, expressed by the following formula: where ε c acpt is the acceptable total strain (adopted based on the operating experience and creep test results)-creep rate for parameters of further service (ε e , Te) provided that the relation ε e ≤ ε c acpt is met.
The method for determination of creep rate for the parameters of further service is shown graphically in Figure 2.
The creep rupture tests with measurement of elongation during test are also used to determine: • The time to end of secondary creep t II and the corresponding elongation ε II The tests are carried out at a constant test temperature T b , which is similar to the operating one, and at stress σ b values higher than those corresponding to the operating one. As a result of each of the tests, the creep curves ε = f (t) at σ b = const are obtained (Figure 3a). The test results are plotted on the diagram in the coordi- Figure 3b). The characteristics obtained in this way allow for determination of the disposable residual life, which is the time of further safe service for the adopted operating parameters.

Short-term creep tests
The disadvantage of the method for determination of residual life using longterm creep tests is the time of waiting for test results, which takes at least 2-3 years. To reduce the duration of these tests and the assessment of residual life, the socalled short-term creep tests with a duration from a few dozen to max 3000-10,000 h are used in the engineering practice. This opens up the possibility for obtaining test results within a maximum of several months and providing good estimation of residual life.
The acceleration of the creep process and the reduction in the duration of tests is obtained in creep tests carried out under uniaxial tension on test specimens taken from the material of a power system component and: • Under constant test stress equal to the operating stress and at different test temperature levels much higher than the operating temperature • At constant test temperature equal to the operating temperature and under different test stress levels much higher than the operating stress T b , test temperature; T r , operating temperature; σ b , test stress; σ r , operating stress; t r , time to rupture; R z av , average temporary creep strength (result of longterm creep tests); t r(r)s , time to rupture for the operating parameters obtained by short-term creep tests as a result of extrapolation; t r(r) , time to rupture for the operating parameters obtained by long-term creep tests of up to 100,000 h.
The manner of presenting results of short-term creep tests conducted under constant stress equal to the operating one (σb = σr = const) for different levels of test temperature Tb and at a constant temperature equal to the operating one (Tb = tr = const) and under different levels of test stress σb as well as the reliability of their estimation is shown graphically in Figures 4 and 5.
Based on at least 40 years' experience of the Creep Test Laboratory of the Institute for Ferrous Metallurgy in comparing results of short-term creep tests and extrapolating them with results of long-term tests of even up to 100,000 h, they were found to be useful for and applicable to the estimation of life and residual life. The determination of residual life for the material after service does not require knowledge of its history and previous operating conditions, and the only requirement is to define the operating parameters of its further service (σ rp , T rp ). Strong convergence of the results obtained in short-and long-term creep tests allows the method of short-term creep tests conducted under constant stress equal to the operating one to be used in practice. A disadvantage of this method is the restriction with regard to the applied test temperature levels arising from precipitation processes that occur in the material. The temperature levels must therefore be selected individually for each grade of material being tested.
Therefore, in the diagnostic practice, only short-term creep tests under constant stress and at different levels of temperature significantly higher than the operating one should be used to determine the residual life of the materials of components operating under creep conditions (pipelines, steam superheater coils). Such tests are recommended for use after exceeding the design service time, which is most often equal to 100,000 h. When the design time is exceeded by 50% (1.5 of t o ), these tests are obligatory.
The reliability of the obtained results of short-term creep tests depends on the requirements for compliance with constant level of preset test temperature T b over the measurement length of the test specimen throughout the duration of the test. The determination of residual life with an error of no more than 20% in relation to the time determined based on long-term creep tests ensures that short-term creep tests are carried out at the test temperature T b (much higher than the operating one) with an accuracy equal to or higher than AE0.5°C over the measurement length of the test specimen, regardless of what the test temperature level is.
3. Method for determination of the time of safe service for material after long-term service under creep conditions beyond the design service time The main step in the proposed method for determination of the time of further safe service is to determine the residual life. Its determination is based on creep test results. These tests have been so far the only known method for determination of service life for actual operating parameters of the materials in use. In creep tests, the decisive factor for their duration is the time to rupture. It cannot be reduced in the development of material characteristics. However, such a possibility exists for the assessment of specific material both in the as-received condition and after service. Nevertheless, the research methods used for this purpose have to be verified with results of long-term creep tests [16]. The reliability of the estimation of time to rupture tr for the operating parameters based on the results of shortterm creep tests at a constant temperature equal to the operating one Tb = Tr = const and under different levels of stress σb higher than the operating one. T b , test temperature; T r , operating temperature; σ b , test stress; σ r , operating stress; t r , time to rupture; R z av , average temporary creep strength (result of long-term creep tests); t z(r)s , time to rupture for the operating parameters obtained in short-term creep tests as a result of extrapolation; t r(r) , time to rupture for the operating parameters obtained in long-term creep tests of up to 100,000 h.
The proposed procedure for determination of residual life and disposable residual life, which is the time of safe service beyond the design time, is presented as the algorithm presented in Figure 6. This algorithm consists of six consecutive steps.
The first step is to select the representative areas for destructive testing. At the designated area in the material of test specimen, the short-term creep tests should be carried out under constant stress on at least three different levels (step 2). Based on the relationships obtained from the results of short-term creep tests, the curves of temporary creep strength at a constant temperature are determined for a few temperature levels (step 3). These creep strength curves allow for determination of the parameter creep strength curve where the parameter is a function of temperature and time to rupture (step 4). The obtained parameter creep resistance curve allows the residual creep strength to be determined for the operating parameters (temperature and stress) of further service (step 5). When the residual creep strength for the parameters of further service is known, the disposable residual life can be determined. It is the safe time of further service for the adopted operating parameters (step 6). The proposed procedure consisting of consecutive steps is presented below on the example of the material after long-term service under creep conditions far beyond the design time.

Selection of the representative areas for destructive testing and of the test material (step 1)
The selection of the representative areas for destructive testing is made after review of the previous service, results of measurements and diagnostic tests carried out on the object, verification calculations and determination of the temperature, and strain and stress distributions by the finite element method for the most strenuous areas [17][18][19].
The test materials selected in this way, with provision of the appropriate access to components for cutting them out and the possibility for repair/replacement, were an elbow and a straight section of the main primary steam pipeline made of 13 HMF steel after approx. 200,000 h service.

Short-term creep tests (step 2)
The short-term creep tests were carried out for six levels of test temperature T b = 600, 620, 640, 650, 660, and 680°C under the constant level of test stress σ b with three different values of σ b = 50, 80, and 100 MPa.
The obtained results of tests on the primary steam pipeline elbow made of 14MoV6-3 steel in the form of logt r = f(T b ) under σ b = const, where t r is the time to rupture in creep test, for three levels of stress σ b1 = 50 MPa, σ b2 = 80 MPa, and σ b3 = 100 MPa are shown in Figure 7.

Temporary creep strength curves at a constant temperature (step 3)
The extrapolation of the obtained relationships of log t r = f(t b ) under σ b = const towards a lower temperature allowed for determination of residual life t re for temperature equal to the extrapolation one.
The results of this extrapolation for one level of temperature T b = const and three levels of stress σ b1 = 50, σ b2 = 80, and σ b3 = 100 MPa allowed the creep strength curves to be plotted in the form of log σ b = f (logt re ) for T b1 = 530, T b2 = 540, T b3 = 560, and T b4 = 580°C, which is presented graphically in Figure 8.

Parameter residual creep strength curve (step 4)
Based on the constant-temperature residual creep strength curves in the form of log σ b = f(logt re ), the residual creep strength t re was determined, and on its basis, the Larson-Miller parameter curve was determined in the form of σ b = f(L-M). Where: L-M = T b (C + log t re ); T b , test temperature; C, material constant, t re , residual creep strength (Figure 9).

Safe time of service beyond the design time (step 5)
In contrast to the actual lifetime equal to the time to rupture of the material subject to creep, which is the residual life t re , the disposable residual life t be is of practical importance. The disposable life is the time until which a structural component can be safely operated under the assumed temperature and stress conditions.
The disposable residual life is part of the residual life. To determine its share in the residual life, defined as the time to the end of the secondary creep t IIe in relation to the time to rupture t re, the rupture creep tests with measurement of elongation during test at a constant temperature T b = 500°C similar to the operating one were carried out.  The results of the creep tests shown as creep curves ε = f(t) at T b = const and the ratio of the disposable residual life t be to the residual life t re depending on the test stress σ b allowed to determine the t be /t re ratio for the stress level equal to the operating stress σ b = σ ep = 50-60 MPa, which is written as t be = 0.55t re (Figure 3).
By knowing the residual life, the disposable residual life can be determined as the time of safe service for the adopted operating parameters. For the material of elbow, the relationships between the disposable residual life t be and the operating temperature of further service T ep for the adopted constant level of operating stress σ ep at the selected levels of operating temperature T ep = 540 and 550°C and under stress σ b = σ ep = 60 and 70 MPa are shown in Figure 10a and for the circumferential welded joint at T ep = 530 and 540°C and operating stress σ ep = 50 MPa in Figure 6, whereas the safe service time of the material of test primary steam pipeline elbow and circumferential welded joint made of 13 HMF steel after approx. 200,000 h service under creep conditions for the operating parameters of further service, which was determined based on the above-mentioned relationships, is summarised in Table 1.

Estimation of exhaustion degree caused by creep (step 6)
The exhaustion degree for materials working under creep conditions was defined in Chapter 3 of the study as the time of service until the time to rupture of this material under the temperature and stress operating conditions (t e /t r ).
For the reviewed case, the residual life t re ( Table 2, column 9) was determined based on the developed relationship of log σ = f(t re ) for the operating temperature of service T e . It was determined for the level of operating stress σ e of the test components as shown in Figure 11.
To determine the operating stress level using the formula where σ e is the operating stress, g act min is the actual minimum wall thickness of component, p e is the operating pressure of component, D Out is the nominal outside diameter, and z is the weakening coefficient, it is necessary to know the specific quantities in the formula. Their designation and values for the test material of elbow and circumferential welded joint are summarised in Table 2, columns 2-7. The table also presents the operating stress σ r values calculated for the test materials after service (column 8).
The life t r is defined as the total time of the previous service t e and residual life t re for the parameters of the previous service (σ e , T e ), that is, t r = t e + t re . Since the exhaustion degree based on the above-mentioned definition can be written as t e /t r and t r = t e + t re , then, in practice, t e /t r = t e /(t e + t re ). Using this relationship, the exhaustion degree ( Table 2,  The most important element necessary to determine the minimum required component wall thickness g oe , which will be able to transfer the required service load (σ ep , T ep ), is to have the characteristics of residual creep strength for the material of test component after service in the form of log σ = f(t re ) for the operating temperature T ep . This is the first step in the adopted procedure.
Step 2 is to determine the residual creep strength R Ze/Tep/tep for the adopted temperature of further service T e and the assumed time of further service t e based on log σ = f(t re ) at T ep = const. The value of R Ze/Tep/tep obtained from the characteristics makes it possible to determine the acceptable stress k for the parameters of further service adopted in this way, which is step 3. To calculate the required minimum wall thickness g oe for the adopted parameters of further service, the nominal outside diameter D Out of the component, the operating pressure p ep , and the construction weakening coefficient z should be defined, which is step 4.
Step 5 is the determination of the required minimum wall thickness g oe of the component for the adopted parameters of further service.
And the last, sixth, step is to compare the obtained value of the calculated required minimum wall thickness g oe to the measured minimum actual thickness g act min . If for the adopted time of further safe service t ep : • g oe < g act min , the component can be allowed for further service.
• g oe ≥ g act min , the component cannot be allowed to continue the operation at the required parameters of further service for the adopted time t ep .
If g oe ≥ g act min and the component cannot be allowed for further service, a shorter time of service should be adopted and the proposed procedure repeated. It may turn out that with a shorter time of further service t ep , for the adopted parameters of further service, the condition of g oe < g act min will be met. This will allow the test component to be operated for such a shorter time t ep .
The proposed procedure in the form of an algorithm consisting of consecutive steps is shown in Figure 12. Its application is presented on the example of the primary steam pipeline elbow and circumferential welded joint made of 13 HMF steel after 200,000 h service.

Characteristics of residual creep strength
The temporary creep strength curve for the material of elbow after 200,000 h service under creep conditions plotted based on results of the tests discussed in Chapter 2 is shown in Figure 13 in the form of lg σ = f(t re ) for the temperature of further service T ep = 540°C.

Determination of residual creep strength for the adopted time of further service
Based on the characteristics in point 2, the residual creep strength at 530 and 540°C for 10,000, 30,000, 50,000, and 100,000 h was determined for the materials of test primary steam pipeline components after 200,000 h service. The obtained results are presented in Table 3.

Determination of the acceptable stress
The value of acceptable stress k is equal to the value of acceptable lower scatter band À20% from the obtained mean value of residual creep strength for the adopted time t ep and temperature T ep . It is calculated from the following formula: The obtained values of acceptable stress k for the adopted levels of temperature T ep = 530, 540°C, and times t e = 10,000, 30,000, 50,000, and 100,000 h of further service are provided in Table 4.

Definition of parameters of further service
The parameters of further service are defined to be the same as the previous ones. To calculate the required minimum wall thickness, in addition to pressure p ep and temperature T ep of the component, it is also necessary to define its nominal outside diameter D Out , nominal wall thickness g n , design pressure p o , operating pressure of further service p e , and construction weakening coefficient z. The required values are summarised in Table 5.

Calculation of the required minimum component wall thickness for the parameters of further service
The calculation of the minimum component wall thickness for the parameters of further service g oe was made based on the following formula: where g oe is the minimum component wall thickness for the parameters of further service, p ep is the operating pressure of further service, D Out is the nominal outside diameter, k is the acceptable stress, and z is the weakening coefficient. The values obtained for pressure p o and p ep and for the adopted different times t ep and temperature levels of further service T ep for the material of test primary steam pipeline elbow made of 13HMF steel are summarised in Table 6.  Table 6. Required minimum wall thickness of primary steam pipeline elbow made of 14MoV6-3 steel for the adopted parameters of further service. Figure 14.
Comparison of the required minimum pipeline elbow wall thickness g oe for the adopted parameters of further service (p ep ,T ep ) to the minimum actual wall thickness g act min measured on the component for the adopted temperature of further service T ep = 540°C.
4.6 Comparison of the calculated required minimum wall thickness to the actual minimum thickness based on measurements taken on the component The obtained results of calculations of the required minimum wall thickness for the parameters of further service ( Table 6) were compared to the actual minimum thickness based on measurements taken on the component ( Table 5, item 3) as shown graphically in Figure 14.
These results meet the condition of g oe < g act min except for the adopted temperature of further service T ep = 540°C and the time of further service t ep = 100,000 h for which this condition is not met and the component could not be allowed for further service of 100,000 h at 540°C for the other adopted operating parameters.