- Email: [email protected]
Roll-expanded plugs for steam generator heating tubes verification of leak tightness over the component lifetime
Nuclear Engineering and Design 263 (2013) 179–186
Contents lists available at SciVerse ScienceDirect
Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes
Roll-expanded plugs for steam generator heating tubes verification of leak tightness over the component lifetime J. Beck ∗ , R. Ziegler, N. Schönheit Areva GmbH, Paul-Gossen-Straße 100, 91058 Erlangen, Germany
h i g h l i g h t s • • • •
Design description of roll-expanded plugs. Experimental simulation of 40 years lifetime of plugged steam generator tubes. Destructive testing for off-design loads. Evaluation of release pressure and tightness before and after the tests.
a r t i c l e
i n f o
Article history: Received 16 August 2012 Received in revised form 11 April 2013 Accepted 16 April 2013
a b s t r a c t Steam generator heating tubes are the boundary between the irradiated primary cycle and the conventional secondary cycle in a pressurized water reactor. Despite their operational task to transfer the heat from the primary to the secondary cycle, these tubes have a crucial safety function: the retention of irradiated primary coolant inside the circuit in all operating, emergency and off-design conditions. The heating tubes are subject to various degradation mechanisms during operation. To verify the integrity of each single tube, nuclear power plants carry out frequent in-service inspections. In case of a tube wall degradation beyond the permissible limit, the tube needs to be taken out of service in order to maintain the overall component integrity. The most common method to do so is to plug a damaged tube by a roll-expanded plug. After plugging, the roll-expanded plug acts as pressure boundary between the primary and the secondary cycle instead of the damaged heating tube. The plug must be able to maintain this function, previously provided by the heating tube, in all operational, emergency and off-design conditions. This article describes the approach to this verification by launching several comprehensive process qualification programmes consisting of mechanical analyses as well as static and dynamic testing programmes. The result was a qualified roll-expanded plug which remains leak-tight even during off-design conditions. © 2013 Elsevier B.V. All rights reserved.
1. Introduction The steam generators (SGs) of pressurized water reactors (PWRs) join the nuclear island with the secondary cycle. The SGs are key components which have a large impact on the plant performance in terms of operation and safety. Since the SGs are the main heat sink and the pressure boundary of the primary cycle, these components are highly safety-related. During operation, SGs are subject to degradation mechanisms which have an impact on the component lifetime. Most effected
Abbreviations: FWLB, feed water line break; LOCA, loss-of-coolant accident; R&D, Research and Development; SG, steam generator. ∗ Corresponding author. Tel.: +55 243 362 9024. E-mail address: [email protected] (J. Beck). 0029-5493/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.nucengdes.2013.04.015
parts are the heating tubes. These transfer the heat and constitute the barrier of the contaminated primary cycle towards the secondary side. Various corrosive attacks caused by deposited impurities on the secondary side as well as low susceptibility of some tube material towards corrosive chemical environments may cause tube wall thinning. Whenever the remaining tube wall has reached the permissible limit, repair measures must be applied. Although various attempts have been made in the past to develop repair mechanisms to keep heating tubes in service, the most common is to remove a tube with a critical wall thinning from service by sealing the inlet and the outlet of the tube with a plug. Different approaches have been made for tube plugging, whereas the preferred solution consists of mechanically expanded plugs inside the tube. Mechanically expanded plugs are differentiated in mandrel-types, using a bolt connected to a mandrel for
180
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Nomenclature Symbols H P p S S T ˛ ε
expansion ratio load pressure permissible load safety factor temperature thermal expansion coefficient strain
Indices G L m p s T TS US
German limit membrane primary secondary tube tube sheet US-American
legislation. The combination of both regulations proved to cover all regulatory demands to date. 2.1.1. ASME Boiler and Pressure Vessel Code Since the year 2000, section XI of the ASME Boiler and Pressure Vessel Code contains a very comprehensive definition of the required compliance of a roll plug in terms of analysis, testing and qualification issues. The ASME Boiler and Pressure Vessel Code is generally accepted by most utilities and national regulatory bodies and – lacking a similar comprehensive definition – ASME is also applicable if similar definite national regulations are missing. Therefore, the ASME requirements were generally followed as an overall guideline for plug design. 2.1.2. German regulatory implications The German regulation does not differentiate in manufacturing and repair activities. Usually, the same standards apply to repair as well as to component manufacture. This implies the verification of leak tightness during all transients specified during component design for the whole lifetime which had to be considered in the testing programme. 2.2. Operational conditions
expansion and mechanical roll-expanded plugs using a rolling tool which is afterwards removed from the SG. Tube repair by mechanical plugs is covered by the ASME Boiler and Pressure Vessel Code Section XI (ASME, 2007). This guideline is very comprehensive and generally applied by most national regulatory bodies. However, national regulations might imply additional obligations on the plug performance and the required qualification of the repair concept. A range of R&D projects in recent years was launched to investigate the different parameters which influence plug tightness and tube repair concepts. The projects aimed at adjusting the plug design to serve most demands of national regulations that sometimes exceed the requirements imposed by ASME. The following paragraphs describe the boundary conditions of a plug used for SG heating tube plugging and explain the design approach taken for the development of a roll-expanded plug. The analyses and the test set-up are described in the following. They resulted in a roll-expanded plug designed to maintain its leak tightness until end-of-life even in beyond-design conditions. 2. Regulatory demands and operating conditions SGs are designed according to the highest nuclear quality standards. Despite the fact that all these standards have the common aim to ensure the highest safety in terms of manufacture and operation, the detailed requirements may differ. An international service provider needs to comply with several of different standards. To minimize the product range a covering approach may be more beneficial, even though the most stringent design requirements need to be applied. Such an approach was taken for the design of the roll-expanded plug. 2.1. Regulatory demands To cover the regulatory requirements of the ASME Boiler and Pressure Vessel Code and the German KTA (Sicherheitstechnische Regel des KTA, KTA 3201.2; Sicherheitstechnische Regel des KTA, KTA 3201.3) as far as applicable, both were considered for the basic design of the mechanical roll plug. This approach combines the comprehensive definition of the plug requirements and the more stringent, although not clearly defined implications of the German
During operation the plug is exposed to varying thermal and mechanical loads resulting from load transients. The plug has to compensate the tube sheet bending during normal operation as well as during emergency conditions. The loading and unloading of the SGs during operation and testing reduces the elasticity of the tube sheet over the life time and has a significant effect on the retention force and the tightness of the plug. In addition to thermal and mechanical fatigue chemical attack needs to be considered: - Potential corrosion attack by primary and secondary side medium. - In addition, no flow at the secondary side with potential impurity concentration and formation of hard deposits. - Surface tension on the roll transition zone on the primary and on secondary side. These factors require a high insusceptibility of the plug material against intergranular attack (IGA) and stress corrosion cracking (SCC). In order to be able to verify the actual condition of a plug in operation these should be inspectable. Space needs to be provided for common eddy-current test (ECT) probes to enter. Besides inspection, removability of the plug is profitable, if repair on the plug or the tube-to-tube sheet connection of the plugged tube is necessary. 3. Design determination The combination of the boundary conditions described above restricts the design of a roll plug to a hollow, thimble-like shape, common to all roll-expanded plugs (see Fig. 1). The plug has a neck on the bottom to facilitate the positioning. The thread in the tip is used for removal. The interface to the heating tube consists of a specific profile with a coating to ensure leak tightness of the expanded plug without any additional measures. 3.1. Determination of the expansion parameters The most sensitive parameter of the plug design is the expansion ratio. This geometrical variable considers the geometry of the tube and of the plug in the expansion region before and after the expansion and defines the degree of plastic deformation.
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
181
Table 1 Geometrical parameters used to determine the expansion ratio. Parameter
Symbol
Expanded tube inner diameter Expanded plug inner diameter Initial plug outer diameter Initial plug inner diameter Wall thickness Clearance between plug and tube before expansion
DT D2 D0 D1 t a
HG =
D2 − D1 − a t
HUS = 1 −
where
DT − D2 D0 − D1
HG = 2 · HUS
a = DT − D0
and
t=
D0 − D1 2
(1)
(2) (3)
In practical applications the different definitions of the expansion ratio may be confusing, so that the formula used shall be verified when comparing expansion ratios of different manufacturers. For SG tube plugging it is not possible to determine the geometry parameters as given in Table 1. Due to high radiation in the primary chamber and since the work is usually carried out on the critical path; a measurement of the geometry is not possible. Therefore, the torque of the rolling tool is recorded, which is directly related to the expansion ratio as long as no other parameter, such as geometry of material properties, is changed. A test series for each material and geometry combination was carried out for the design of the roll-expanded plugs to determine the optimal expansion ratio or torque setting prior to the verification described herein. 3.2. Material selection Fig. 1. Plug geometry.
Ideally, the plug and tube are expanded up to the tensile strength of the tube sheet material to achieve the highest retention force (Schweigerer, 1978). Due to the complex interaction of different parameters it is practicable to determine the required expansion ratio in test series. This approach is generally recommended by mandrel manufacturers. There are several definitions of the expansion ratio, although all are based on the initial geometry before roll expansion and the residual geometry afterwards. Comparing the expansion ratio of different origins, it is necessary to define the calculation method. 1. The definition of the German DIN standard DIN 28187 (Normenausschuss Chemischer Apparatebau im DIN: DIN 28187) relates to the areas in the plane of the expansion area before and after expansion. The result is a small value where large differences of the transmitted forces result only in a very small difference in the value of the expansion ratio. In practice, these values proved to be difficult to handle. 2. In German SG design the expansion ratio HG is related to the diameter as expressed in formula (1) below. This results in larger values in the expansion ratio and differences appear to be more obvious. This definition of the expansion ratio is used here. 3. In the US, the expansion ratio HUS is related to the radius and the definition according to formula (2) is used. As Eq. (1) is related to the diameter and the other to the radius, the conversion factor is 2 (Eq. (3)). This factor is derived by inserting the values for a and t in the formula for HG .
Due to the high corrosion resistance Alloy 690 (SB 166 UNS 006690) was chosen as plug material: To enhance the sealing capabilities of the expanded area, a specific interface was created using a coating as sealing agent. To verify the corrosion resistance of this combination, corrosion tests were applied prior to the verification described herein. 3.3. Stress analysis The stress analysis of the roll-expanded plug was carried out according to ASME Boiler and Vessel Code Section III for primary and secondary stresses while taking all relevant load cases into account. Primary stresses result from the pressure load of the plug. Secondary stresses are caused by tube sheet deflection. As the retention force of the plug depends on the elasticity of the tube sheet in the expansion area this deflection needs to be considered for this calculation. 3.3.1. Analysis model The models used for the analysis of primary and secondary stresses are shown in Fig. 2(a)) shows the boundary conditions of the model considering the assembly plug/tube–tube sheet. The analysis of primary stresses was carried out by means of model (b). A uniform temperature distribution was assumed, which is a permissible assumption since the plug is close to the primary fluid and the primary transients of concern have no significant influence on the temperature distribution at this location. Therefore, only primary and secondary side pressure was considered. The pressure was applied as shown in Fig. 2(c)).
182
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Fig. 2. FE model used for the analysis: (a) boundary conditions of the assembly; (b) model used for the analysis of primary stresses; (c) pressure applied for the analysis of primary stresses; and (d) loads for secondary stresses in combination with tube sheet deflection.
The inner surface of the plug and the outer surface below the expansion area were loaded with primary pressure while secondary side pressure was applied to the outer surface above the expansion area. For the analysis of secondary stresses models (c) and (d) were combined. Temperature again was assumed to be homogeneous. However, temperature and pressure loads which result in a tube sheet deflection were calculated and applied as boundary conditions.
3.3.2. Tube sheet deflection In a sub-model the maximum displacement of the tube sheet was determined in radial direction. This deflection was considered in the analysis of the secondary stresses as shown in Fig. 3. In a covering approach the largest deflection in the middle of the tube sheet was considered. The divider plate was neglected, which is a conservative assumption yielding larger deflections. From the relevant transients the maximum horizontal displacement x
was calculated and applied as a boundary condition in the finite element model. 3.3.3. Results of the stress analysis The analysis gave the highest utilization for design loads for limit Pm < Sm and PL < 1.5 Sm which showed an utilization of 77.7%. All other utilization factors were almost significantly below that value. Fig. 4 shows the graphical output of the results. Apparently, the expansion area is the zone with the highest stresses whereas the stresses in the remaining sections of the plug are low. 4. Experimental verification An experimental approach was taken to verify the suitability of the plug design. The mechanical parameters of the plugs were examined in static and dynamic pressure tests. First, different operational tube conditions were hydrostatically tested. This limited the
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
183
amount of tests to be conducted. To verify the plug performance at the end of life, dynamic tests were carried out. 4.1. Hydrostatic testing By hydrostatic tests the plug release pressure and the tightness were tested in ambient conditions in differently pre-conditioned heating tubes. The tube conditions were • Clean/as delivered. • Corroded/oxidized. • Corroded/oxidized with debris. It was found that the release pressure for the tested plug design was lowest for clean/as delivered tubes. The corroded/oxidized surface increased the friction between the roll-expanded plug and the test tube, yielding higher plug release forces. The minimum release pressure of a plug for 19.05 × 1.09 mm tubing was 56.0 MPa. The static release pressure was significantly above the ASME criterion of S > 3.0 for normal operation and S > 1.43 for accident conditions. The maximum Helium leak rate of 1.8 × 10−7 cm3 /s/bar was significantly below the permissible leak rate of 1 × 10−6 cm3 /s/bar However, these margins were necessary since these relate to newly set plugs. The aim of the plug tests was to achieve the required values according to the component lifetime which was determined by dynamic tests. 4.2. Dynamic testing
Fig. 3. Consideration of the tube sheet deflection.
The dynamic test programme simulated pressure and temperature gradients inside the tubes superimposed by sheet deflection. Due to bending and deflection of the tube sheet during different load transients, the elasticity of the tube-to-tube sheet connection as well as the plug-to-tube connection decreases. The aim of these tests was to determine the tightness and the release
Fig. 4. Analysis results for design load conditions.
184
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Fig. 5. Conversion of the measured strain according to Schweigerer (1978).
pressure caused by the roll-expanded plugs at end of life including emergency transients and off-design loads. 4.2.1. Dynamic test set-up For the set up, the relevant covering load transients were determined and the mean ligament surface strain on the primary and secondary side of the tube sheet was calculated. This strain was converted to a tube sheet mock up of original qualified material but only of 150 mm in thickness. Furthermore, an addition in strain was applied according to formula (4) made to consider the different temperature expansion coefficients of the materials. εtemp = (˛plug − ˛TS ) · T
(4)
where εtemp is the additional strain due to the different expansion coefficients, ˛plug and ˛TS are the thermal expansion coefficients of the plug or of the tube sheet material and T is the temperature difference related to ambient temperature. The thermal expansion of the tube was neglected since the SG heating tubes of U-tube heat exchangers are commonly made of Alloy 600, Alloy 690 or of Alloy 800 which all have a higher thermal expansion coefficient than the tube sheet material made of ferritic steel such as SA 508 class 2 which would increase the retention force. The strain caused by tube sheet deflection was to be measured by strain gauges. Due to the small ligaments of the tube sheet, the strain gauges were placed in the middle of the ligaments. Therefore, a factor was applied according to Schweigerer (1978) to convert the
Fig. 7. Test set-up with test specimen and pressure lines.
ligament strain measured in the middle of the ligament to the mean surface ligament strain as shown in Fig. 5. The strain was applied by a stress–strain measuring. Therefore, the tube sheet mock up with an outer diameter of 1400 mm was slotted in order to realize the desired deflection. On the outer diameter it was fixed with clamps and on the circumference of the test field with the specimen of a loading ring fixed by screws. A schematic overview of this set up is given in Fig. 6. Altogether four different plug geometries and material combinations were tested. Five specimens per set were used. Each specimen was connected to a pressure line on both sides of the plug to superimpose the tube sheet deflection with the corresponding primary and secondary side pressure of the tested transients. The pressure of each line was recorded simultaneously to detect any pressure drop which may indicate a leakage during the tests. The test set up is shown in Fig. 7.
4.2.2. Dynamic test sequence and destructive testing After completion of the test set-up the Helium leak tightness was determined. The following dynamic testing consisted of two parts.
Fig. 6. Overview of the set-up for dynamic testing.
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
185
Fig. 8. Periods for cyclic testing; one period corresponds to 10 years lifetime.
First, the relevant transients were tested by cyclic testing to simulate a 40 years lifetime. The test sequence consisted of four periods of Levels A and B loads. In order to minimize the amount of tests, only loads with fatigue relevance for the tube sheet were considered. For these, the maximum strain for each relevant load case was determined. For testing, transients with large strain on the primary side (negative values) side were applied alternately with such having a large strain on secondary side (positive values) as shown in Fig. 8. Each testing period corresponded to a 10 years lifetime. After the sequence of four testing periods, Levels C and D loads relevant for the tube sheet deflection were tested for one period. Table 2 and Fig. 8 illustrate the testing sequences. Fig. 9 shows the destructive testing after the cyclic tests. At first, a tightness test (2) was performed. Then, a high strain on the secondary side was applied to minimize the plug retention force followed by a tightness test at 17.5 MPa (4), another high secondary side strain combined with a tightness test at 10.3 MPa (5) and a further tightness test at 17.5 MPa (6).
After destructive testing the specimen were once more tested for Helium leakage (7). Subsequently, the remaining release pressure of the test specimen was determined by application of hydrostatic pressure (8). 4.2.3. Dynamic test results The minimum release pressure during hydrostatic pressure tests performed after the dynamic testing was 47.0 MPa for 19.05 × 1.09 mm tubing. None of the plugs showed any significant leakage. However, the minimum release pressure of the plugs decreased from 56.0 MPa before cyclic testing to 47.0 MPa after the cyclic tests. The plug release pressure of 47.0 MPa was compared to highest pressure difference of 19.0 MPa which occurs during feed water line break in accident conditions and to the pressure of 10.5 MPa during normal operation. pTS in accident conditions = 19.0 MPa (best estimate value) pTS min tested = 47.0 MPa S = 2.47 > 1.43
Table 2 Load cases for cyclic testing. Relevant transients
Test strain et −3
Levels A and B (normal operation and upset conditions)
Primary side hydrostatic test – in service Secondary side hydrostatic test – in service Heat-up and cool-down (Level A) Primary side leak test Tube leak test (c) Tube leak test (d)
−1.57 × 10 1.35 × 10−3 −1.73 × 10−3 −1.63 × 10−3 0.52 × 10−3 0.75 × 10−3
Level C (emergency conditions)
Small steam line break Complete loss of flow
Covered by Level D
Level D (faulted conditions)
Design basis loss of coolant accident (LOCA) Main steam line break (MSLB) Feed water line break (FWLB) Reactor coolant pump locked rotor
1.17 × 10−3 −2.14 × 10−3 −2.28 × 10−3 −1.34 × 10−3
Test cycles NB 1 1 130 39 39 130
1 7 1 6
186
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Fig. 9. Destructive testing following the cyclic tests.
pTS in normal operation = 10.5 MPa (best estimate value) pTS min tested = 47.0 MPa S = 4.47 > 3.0 The comparison shows that even after loading with beyonddesign deflection, the plugs fulfil the ASME criteria for accident and normal operation conditions. The remaining margin of the plugs is still considerable and the expansion parameters were confirmed. 5. Conclusion To meet the different legislative and regulatory boundary conditions of utilities in different countries, AREVA developed a roll plug for the repair of SG heating tubes which covers the majority of requirements set forth by current standards. Due to the covering approach taken it was not always possible to follow a single regulation since none covered all relevant aspects. Where necessary, adaptations where made to limit testing and cover more stringent regulations. The result was a comprehensive testing programme simulating real life conditions of SG operation with special focus
on the operational implications of the plug performance. With plug design and plugging parameters, even beyond-design cases were tested while the acceptance criterion in terms of tightness and release pressure continued to be fulfilled. References American Society of Mechanical Engineers (ASME), 2007. Boiler and Pressure Vessel Code, Edition 2007 without Addendum Section XI: Rules for Inservice Inspection of Nuclear Power Plant Components: Sub-Section IWA 4713 Heat Exchanger Tube Plugging by Expansion. Normenausschuss Chemischer Apparatebau im DIN: DIN 28187: RohrbündelWärmeaustauscher – Rohr/Rohrboden Befestigungen//Tubular heat exchangers – connections tube to tube sheet, September 2009. Sicherheitstechnische Regel des KTA, KTA 3201.2: Komponenten des Primärkreises von Leichtwasserreaktoren, Teil 2: Auslegung, Konstruktion und Berechnung, Fassung 6/96. Sicherheitstechnische Regel des KTA, KTA 3201.3: Komponenten des Primärkreises von Leichtwasserreaktoren, Teil 3: Herstellung, Fassung 11/07. Schweigerer, S., 1978. Festigkeitsberechnung im Dampfkessel-, Behälter-und Rohrleitungsbau 3. Auflage, Springer Verlag, Heidelberg.
Contents lists available at SciVerse ScienceDirect
Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes
Roll-expanded plugs for steam generator heating tubes verification of leak tightness over the component lifetime J. Beck ∗ , R. Ziegler, N. Schönheit Areva GmbH, Paul-Gossen-Straße 100, 91058 Erlangen, Germany
h i g h l i g h t s • • • •
Design description of roll-expanded plugs. Experimental simulation of 40 years lifetime of plugged steam generator tubes. Destructive testing for off-design loads. Evaluation of release pressure and tightness before and after the tests.
a r t i c l e
i n f o
Article history: Received 16 August 2012 Received in revised form 11 April 2013 Accepted 16 April 2013
a b s t r a c t Steam generator heating tubes are the boundary between the irradiated primary cycle and the conventional secondary cycle in a pressurized water reactor. Despite their operational task to transfer the heat from the primary to the secondary cycle, these tubes have a crucial safety function: the retention of irradiated primary coolant inside the circuit in all operating, emergency and off-design conditions. The heating tubes are subject to various degradation mechanisms during operation. To verify the integrity of each single tube, nuclear power plants carry out frequent in-service inspections. In case of a tube wall degradation beyond the permissible limit, the tube needs to be taken out of service in order to maintain the overall component integrity. The most common method to do so is to plug a damaged tube by a roll-expanded plug. After plugging, the roll-expanded plug acts as pressure boundary between the primary and the secondary cycle instead of the damaged heating tube. The plug must be able to maintain this function, previously provided by the heating tube, in all operational, emergency and off-design conditions. This article describes the approach to this verification by launching several comprehensive process qualification programmes consisting of mechanical analyses as well as static and dynamic testing programmes. The result was a qualified roll-expanded plug which remains leak-tight even during off-design conditions. © 2013 Elsevier B.V. All rights reserved.
1. Introduction The steam generators (SGs) of pressurized water reactors (PWRs) join the nuclear island with the secondary cycle. The SGs are key components which have a large impact on the plant performance in terms of operation and safety. Since the SGs are the main heat sink and the pressure boundary of the primary cycle, these components are highly safety-related. During operation, SGs are subject to degradation mechanisms which have an impact on the component lifetime. Most effected
Abbreviations: FWLB, feed water line break; LOCA, loss-of-coolant accident; R&D, Research and Development; SG, steam generator. ∗ Corresponding author. Tel.: +55 243 362 9024. E-mail address: [email protected] (J. Beck). 0029-5493/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.nucengdes.2013.04.015
parts are the heating tubes. These transfer the heat and constitute the barrier of the contaminated primary cycle towards the secondary side. Various corrosive attacks caused by deposited impurities on the secondary side as well as low susceptibility of some tube material towards corrosive chemical environments may cause tube wall thinning. Whenever the remaining tube wall has reached the permissible limit, repair measures must be applied. Although various attempts have been made in the past to develop repair mechanisms to keep heating tubes in service, the most common is to remove a tube with a critical wall thinning from service by sealing the inlet and the outlet of the tube with a plug. Different approaches have been made for tube plugging, whereas the preferred solution consists of mechanically expanded plugs inside the tube. Mechanically expanded plugs are differentiated in mandrel-types, using a bolt connected to a mandrel for
180
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Nomenclature Symbols H P p S S T ˛ ε
expansion ratio load pressure permissible load safety factor temperature thermal expansion coefficient strain
Indices G L m p s T TS US
German limit membrane primary secondary tube tube sheet US-American
legislation. The combination of both regulations proved to cover all regulatory demands to date. 2.1.1. ASME Boiler and Pressure Vessel Code Since the year 2000, section XI of the ASME Boiler and Pressure Vessel Code contains a very comprehensive definition of the required compliance of a roll plug in terms of analysis, testing and qualification issues. The ASME Boiler and Pressure Vessel Code is generally accepted by most utilities and national regulatory bodies and – lacking a similar comprehensive definition – ASME is also applicable if similar definite national regulations are missing. Therefore, the ASME requirements were generally followed as an overall guideline for plug design. 2.1.2. German regulatory implications The German regulation does not differentiate in manufacturing and repair activities. Usually, the same standards apply to repair as well as to component manufacture. This implies the verification of leak tightness during all transients specified during component design for the whole lifetime which had to be considered in the testing programme. 2.2. Operational conditions
expansion and mechanical roll-expanded plugs using a rolling tool which is afterwards removed from the SG. Tube repair by mechanical plugs is covered by the ASME Boiler and Pressure Vessel Code Section XI (ASME, 2007). This guideline is very comprehensive and generally applied by most national regulatory bodies. However, national regulations might imply additional obligations on the plug performance and the required qualification of the repair concept. A range of R&D projects in recent years was launched to investigate the different parameters which influence plug tightness and tube repair concepts. The projects aimed at adjusting the plug design to serve most demands of national regulations that sometimes exceed the requirements imposed by ASME. The following paragraphs describe the boundary conditions of a plug used for SG heating tube plugging and explain the design approach taken for the development of a roll-expanded plug. The analyses and the test set-up are described in the following. They resulted in a roll-expanded plug designed to maintain its leak tightness until end-of-life even in beyond-design conditions. 2. Regulatory demands and operating conditions SGs are designed according to the highest nuclear quality standards. Despite the fact that all these standards have the common aim to ensure the highest safety in terms of manufacture and operation, the detailed requirements may differ. An international service provider needs to comply with several of different standards. To minimize the product range a covering approach may be more beneficial, even though the most stringent design requirements need to be applied. Such an approach was taken for the design of the roll-expanded plug. 2.1. Regulatory demands To cover the regulatory requirements of the ASME Boiler and Pressure Vessel Code and the German KTA (Sicherheitstechnische Regel des KTA, KTA 3201.2; Sicherheitstechnische Regel des KTA, KTA 3201.3) as far as applicable, both were considered for the basic design of the mechanical roll plug. This approach combines the comprehensive definition of the plug requirements and the more stringent, although not clearly defined implications of the German
During operation the plug is exposed to varying thermal and mechanical loads resulting from load transients. The plug has to compensate the tube sheet bending during normal operation as well as during emergency conditions. The loading and unloading of the SGs during operation and testing reduces the elasticity of the tube sheet over the life time and has a significant effect on the retention force and the tightness of the plug. In addition to thermal and mechanical fatigue chemical attack needs to be considered: - Potential corrosion attack by primary and secondary side medium. - In addition, no flow at the secondary side with potential impurity concentration and formation of hard deposits. - Surface tension on the roll transition zone on the primary and on secondary side. These factors require a high insusceptibility of the plug material against intergranular attack (IGA) and stress corrosion cracking (SCC). In order to be able to verify the actual condition of a plug in operation these should be inspectable. Space needs to be provided for common eddy-current test (ECT) probes to enter. Besides inspection, removability of the plug is profitable, if repair on the plug or the tube-to-tube sheet connection of the plugged tube is necessary. 3. Design determination The combination of the boundary conditions described above restricts the design of a roll plug to a hollow, thimble-like shape, common to all roll-expanded plugs (see Fig. 1). The plug has a neck on the bottom to facilitate the positioning. The thread in the tip is used for removal. The interface to the heating tube consists of a specific profile with a coating to ensure leak tightness of the expanded plug without any additional measures. 3.1. Determination of the expansion parameters The most sensitive parameter of the plug design is the expansion ratio. This geometrical variable considers the geometry of the tube and of the plug in the expansion region before and after the expansion and defines the degree of plastic deformation.
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
181
Table 1 Geometrical parameters used to determine the expansion ratio. Parameter
Symbol
Expanded tube inner diameter Expanded plug inner diameter Initial plug outer diameter Initial plug inner diameter Wall thickness Clearance between plug and tube before expansion
DT D2 D0 D1 t a
HG =
D2 − D1 − a t
HUS = 1 −
where
DT − D2 D0 − D1
HG = 2 · HUS
a = DT − D0
and
t=
D0 − D1 2
(1)
(2) (3)
In practical applications the different definitions of the expansion ratio may be confusing, so that the formula used shall be verified when comparing expansion ratios of different manufacturers. For SG tube plugging it is not possible to determine the geometry parameters as given in Table 1. Due to high radiation in the primary chamber and since the work is usually carried out on the critical path; a measurement of the geometry is not possible. Therefore, the torque of the rolling tool is recorded, which is directly related to the expansion ratio as long as no other parameter, such as geometry of material properties, is changed. A test series for each material and geometry combination was carried out for the design of the roll-expanded plugs to determine the optimal expansion ratio or torque setting prior to the verification described herein. 3.2. Material selection Fig. 1. Plug geometry.
Ideally, the plug and tube are expanded up to the tensile strength of the tube sheet material to achieve the highest retention force (Schweigerer, 1978). Due to the complex interaction of different parameters it is practicable to determine the required expansion ratio in test series. This approach is generally recommended by mandrel manufacturers. There are several definitions of the expansion ratio, although all are based on the initial geometry before roll expansion and the residual geometry afterwards. Comparing the expansion ratio of different origins, it is necessary to define the calculation method. 1. The definition of the German DIN standard DIN 28187 (Normenausschuss Chemischer Apparatebau im DIN: DIN 28187) relates to the areas in the plane of the expansion area before and after expansion. The result is a small value where large differences of the transmitted forces result only in a very small difference in the value of the expansion ratio. In practice, these values proved to be difficult to handle. 2. In German SG design the expansion ratio HG is related to the diameter as expressed in formula (1) below. This results in larger values in the expansion ratio and differences appear to be more obvious. This definition of the expansion ratio is used here. 3. In the US, the expansion ratio HUS is related to the radius and the definition according to formula (2) is used. As Eq. (1) is related to the diameter and the other to the radius, the conversion factor is 2 (Eq. (3)). This factor is derived by inserting the values for a and t in the formula for HG .
Due to the high corrosion resistance Alloy 690 (SB 166 UNS 006690) was chosen as plug material: To enhance the sealing capabilities of the expanded area, a specific interface was created using a coating as sealing agent. To verify the corrosion resistance of this combination, corrosion tests were applied prior to the verification described herein. 3.3. Stress analysis The stress analysis of the roll-expanded plug was carried out according to ASME Boiler and Vessel Code Section III for primary and secondary stresses while taking all relevant load cases into account. Primary stresses result from the pressure load of the plug. Secondary stresses are caused by tube sheet deflection. As the retention force of the plug depends on the elasticity of the tube sheet in the expansion area this deflection needs to be considered for this calculation. 3.3.1. Analysis model The models used for the analysis of primary and secondary stresses are shown in Fig. 2(a)) shows the boundary conditions of the model considering the assembly plug/tube–tube sheet. The analysis of primary stresses was carried out by means of model (b). A uniform temperature distribution was assumed, which is a permissible assumption since the plug is close to the primary fluid and the primary transients of concern have no significant influence on the temperature distribution at this location. Therefore, only primary and secondary side pressure was considered. The pressure was applied as shown in Fig. 2(c)).
182
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Fig. 2. FE model used for the analysis: (a) boundary conditions of the assembly; (b) model used for the analysis of primary stresses; (c) pressure applied for the analysis of primary stresses; and (d) loads for secondary stresses in combination with tube sheet deflection.
The inner surface of the plug and the outer surface below the expansion area were loaded with primary pressure while secondary side pressure was applied to the outer surface above the expansion area. For the analysis of secondary stresses models (c) and (d) were combined. Temperature again was assumed to be homogeneous. However, temperature and pressure loads which result in a tube sheet deflection were calculated and applied as boundary conditions.
3.3.2. Tube sheet deflection In a sub-model the maximum displacement of the tube sheet was determined in radial direction. This deflection was considered in the analysis of the secondary stresses as shown in Fig. 3. In a covering approach the largest deflection in the middle of the tube sheet was considered. The divider plate was neglected, which is a conservative assumption yielding larger deflections. From the relevant transients the maximum horizontal displacement x
was calculated and applied as a boundary condition in the finite element model. 3.3.3. Results of the stress analysis The analysis gave the highest utilization for design loads for limit Pm < Sm and PL < 1.5 Sm which showed an utilization of 77.7%. All other utilization factors were almost significantly below that value. Fig. 4 shows the graphical output of the results. Apparently, the expansion area is the zone with the highest stresses whereas the stresses in the remaining sections of the plug are low. 4. Experimental verification An experimental approach was taken to verify the suitability of the plug design. The mechanical parameters of the plugs were examined in static and dynamic pressure tests. First, different operational tube conditions were hydrostatically tested. This limited the
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
183
amount of tests to be conducted. To verify the plug performance at the end of life, dynamic tests were carried out. 4.1. Hydrostatic testing By hydrostatic tests the plug release pressure and the tightness were tested in ambient conditions in differently pre-conditioned heating tubes. The tube conditions were • Clean/as delivered. • Corroded/oxidized. • Corroded/oxidized with debris. It was found that the release pressure for the tested plug design was lowest for clean/as delivered tubes. The corroded/oxidized surface increased the friction between the roll-expanded plug and the test tube, yielding higher plug release forces. The minimum release pressure of a plug for 19.05 × 1.09 mm tubing was 56.0 MPa. The static release pressure was significantly above the ASME criterion of S > 3.0 for normal operation and S > 1.43 for accident conditions. The maximum Helium leak rate of 1.8 × 10−7 cm3 /s/bar was significantly below the permissible leak rate of 1 × 10−6 cm3 /s/bar However, these margins were necessary since these relate to newly set plugs. The aim of the plug tests was to achieve the required values according to the component lifetime which was determined by dynamic tests. 4.2. Dynamic testing
Fig. 3. Consideration of the tube sheet deflection.
The dynamic test programme simulated pressure and temperature gradients inside the tubes superimposed by sheet deflection. Due to bending and deflection of the tube sheet during different load transients, the elasticity of the tube-to-tube sheet connection as well as the plug-to-tube connection decreases. The aim of these tests was to determine the tightness and the release
Fig. 4. Analysis results for design load conditions.
184
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Fig. 5. Conversion of the measured strain according to Schweigerer (1978).
pressure caused by the roll-expanded plugs at end of life including emergency transients and off-design loads. 4.2.1. Dynamic test set-up For the set up, the relevant covering load transients were determined and the mean ligament surface strain on the primary and secondary side of the tube sheet was calculated. This strain was converted to a tube sheet mock up of original qualified material but only of 150 mm in thickness. Furthermore, an addition in strain was applied according to formula (4) made to consider the different temperature expansion coefficients of the materials. εtemp = (˛plug − ˛TS ) · T
(4)
where εtemp is the additional strain due to the different expansion coefficients, ˛plug and ˛TS are the thermal expansion coefficients of the plug or of the tube sheet material and T is the temperature difference related to ambient temperature. The thermal expansion of the tube was neglected since the SG heating tubes of U-tube heat exchangers are commonly made of Alloy 600, Alloy 690 or of Alloy 800 which all have a higher thermal expansion coefficient than the tube sheet material made of ferritic steel such as SA 508 class 2 which would increase the retention force. The strain caused by tube sheet deflection was to be measured by strain gauges. Due to the small ligaments of the tube sheet, the strain gauges were placed in the middle of the ligaments. Therefore, a factor was applied according to Schweigerer (1978) to convert the
Fig. 7. Test set-up with test specimen and pressure lines.
ligament strain measured in the middle of the ligament to the mean surface ligament strain as shown in Fig. 5. The strain was applied by a stress–strain measuring. Therefore, the tube sheet mock up with an outer diameter of 1400 mm was slotted in order to realize the desired deflection. On the outer diameter it was fixed with clamps and on the circumference of the test field with the specimen of a loading ring fixed by screws. A schematic overview of this set up is given in Fig. 6. Altogether four different plug geometries and material combinations were tested. Five specimens per set were used. Each specimen was connected to a pressure line on both sides of the plug to superimpose the tube sheet deflection with the corresponding primary and secondary side pressure of the tested transients. The pressure of each line was recorded simultaneously to detect any pressure drop which may indicate a leakage during the tests. The test set up is shown in Fig. 7.
4.2.2. Dynamic test sequence and destructive testing After completion of the test set-up the Helium leak tightness was determined. The following dynamic testing consisted of two parts.
Fig. 6. Overview of the set-up for dynamic testing.
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
185
Fig. 8. Periods for cyclic testing; one period corresponds to 10 years lifetime.
First, the relevant transients were tested by cyclic testing to simulate a 40 years lifetime. The test sequence consisted of four periods of Levels A and B loads. In order to minimize the amount of tests, only loads with fatigue relevance for the tube sheet were considered. For these, the maximum strain for each relevant load case was determined. For testing, transients with large strain on the primary side (negative values) side were applied alternately with such having a large strain on secondary side (positive values) as shown in Fig. 8. Each testing period corresponded to a 10 years lifetime. After the sequence of four testing periods, Levels C and D loads relevant for the tube sheet deflection were tested for one period. Table 2 and Fig. 8 illustrate the testing sequences. Fig. 9 shows the destructive testing after the cyclic tests. At first, a tightness test (2) was performed. Then, a high strain on the secondary side was applied to minimize the plug retention force followed by a tightness test at 17.5 MPa (4), another high secondary side strain combined with a tightness test at 10.3 MPa (5) and a further tightness test at 17.5 MPa (6).
After destructive testing the specimen were once more tested for Helium leakage (7). Subsequently, the remaining release pressure of the test specimen was determined by application of hydrostatic pressure (8). 4.2.3. Dynamic test results The minimum release pressure during hydrostatic pressure tests performed after the dynamic testing was 47.0 MPa for 19.05 × 1.09 mm tubing. None of the plugs showed any significant leakage. However, the minimum release pressure of the plugs decreased from 56.0 MPa before cyclic testing to 47.0 MPa after the cyclic tests. The plug release pressure of 47.0 MPa was compared to highest pressure difference of 19.0 MPa which occurs during feed water line break in accident conditions and to the pressure of 10.5 MPa during normal operation. pTS in accident conditions = 19.0 MPa (best estimate value) pTS min tested = 47.0 MPa S = 2.47 > 1.43
Table 2 Load cases for cyclic testing. Relevant transients
Test strain et −3
Levels A and B (normal operation and upset conditions)
Primary side hydrostatic test – in service Secondary side hydrostatic test – in service Heat-up and cool-down (Level A) Primary side leak test Tube leak test (c) Tube leak test (d)
−1.57 × 10 1.35 × 10−3 −1.73 × 10−3 −1.63 × 10−3 0.52 × 10−3 0.75 × 10−3
Level C (emergency conditions)
Small steam line break Complete loss of flow
Covered by Level D
Level D (faulted conditions)
Design basis loss of coolant accident (LOCA) Main steam line break (MSLB) Feed water line break (FWLB) Reactor coolant pump locked rotor
1.17 × 10−3 −2.14 × 10−3 −2.28 × 10−3 −1.34 × 10−3
Test cycles NB 1 1 130 39 39 130
1 7 1 6
186
J. Beck et al. / Nuclear Engineering and Design 263 (2013) 179–186
Fig. 9. Destructive testing following the cyclic tests.
pTS in normal operation = 10.5 MPa (best estimate value) pTS min tested = 47.0 MPa S = 4.47 > 3.0 The comparison shows that even after loading with beyonddesign deflection, the plugs fulfil the ASME criteria for accident and normal operation conditions. The remaining margin of the plugs is still considerable and the expansion parameters were confirmed. 5. Conclusion To meet the different legislative and regulatory boundary conditions of utilities in different countries, AREVA developed a roll plug for the repair of SG heating tubes which covers the majority of requirements set forth by current standards. Due to the covering approach taken it was not always possible to follow a single regulation since none covered all relevant aspects. Where necessary, adaptations where made to limit testing and cover more stringent regulations. The result was a comprehensive testing programme simulating real life conditions of SG operation with special focus
on the operational implications of the plug performance. With plug design and plugging parameters, even beyond-design cases were tested while the acceptance criterion in terms of tightness and release pressure continued to be fulfilled. References American Society of Mechanical Engineers (ASME), 2007. Boiler and Pressure Vessel Code, Edition 2007 without Addendum Section XI: Rules for Inservice Inspection of Nuclear Power Plant Components: Sub-Section IWA 4713 Heat Exchanger Tube Plugging by Expansion. Normenausschuss Chemischer Apparatebau im DIN: DIN 28187: RohrbündelWärmeaustauscher – Rohr/Rohrboden Befestigungen//Tubular heat exchangers – connections tube to tube sheet, September 2009. Sicherheitstechnische Regel des KTA, KTA 3201.2: Komponenten des Primärkreises von Leichtwasserreaktoren, Teil 2: Auslegung, Konstruktion und Berechnung, Fassung 6/96. Sicherheitstechnische Regel des KTA, KTA 3201.3: Komponenten des Primärkreises von Leichtwasserreaktoren, Teil 3: Herstellung, Fassung 11/07. Schweigerer, S., 1978. Festigkeitsberechnung im Dampfkessel-, Behälter-und Rohrleitungsbau 3. Auflage, Springer Verlag, Heidelberg.