Towards an improved understanding of the mechanisms involved in the - - PowerPoint PPT Presentation

WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN

Towards an improved understanding of the mechanisms involved in the increased hydrogen uptake and corrosion at high burnups in zirconium based claddings

Sousan Abolhassani:: Senior Scientist :: Paul Scherrer Institut ASTM International, Manchester, 21st May 2019

WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN

Sousan Abolhassani1, Adrienn Baris1, Robin Grabherr1, Jonathan Hawes1, Aaron Colldeweih1, Radovan Vanta1, Renato Restani1, Armin Hermann1, Johannes Bertsch1, Melanie Chollet1, Goutam Kuri1, Matthias Martin1, Stephane Portier1, Holger Wiese1, Herbert Schweikert1, Gerhard Bart1, Katja Ammon2, Guido Ledergerber2, Magnus Limbäck3

1Laboratory for Nuclear Materials, Nuclear Fuels Group, NES, and AHL, NES, Paul Scherrer Institut, 5232 Villigen PSI, Switzerland. 2Kernkraftwerk Leibstadt AG, CH-5325 Leibstadt, Switzerland, 3 Westinghouse Electric Sweden AB, SE-72163 Västerås, Sweden.

The aim of the project

• To search for the causes of increased H Uptake at high burnups
• The strategy selected:

− Step 1: To study a very high burnup cladding (here a 9-cycle BWR) and look at the hints pointing to changes in characteristics. − Step-2: Then extrapolate the changes to lower burnup examples of the same cladding grade under very similar conditions − Step-3: Determine the indicators − Step-4: Check if they are valid in other alloys or other reactors

• Evaluate which regime is acting at which stage?

Page 4

Hydrogen content as a function of oxide thickness

200 400 600 800 20 40 60 80 100 120 140

Average oxide thickness (micrometer) Hydrogen content (ppm)

LK3/L 6 c LK3/L 7 c LK3/L 9 c LK3/L 5 c LK3/L 3 c LK2/L 7 c LK2/L 6 c LK2/L 3 c LK2/L 6 c new

Different alloys (LK3/L and LK2/L cladding) behave differently. For identical elevation hydrogen uptake increases with the burnup, the hydrogen pickup fraction (HPF) is not constant and it increases with the high burnup in one alloy and decreases in a second

• alloy. Data in graph is for mid-span peak burn-up elevation, approx. 2000 mm.

[Abolhassani et al. ASTM 17th, 2013].

Parameters influencing hydrogen distribution

Correlation between oxidation and hydrogen content (example of a BWR)

100 % 50 % 25 % 15 %

HPF

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

6

Comp. Specimen Sn Fe Cr Ni Nb Fe+Cr(+Ni) C ppm O ppm Si ppm H.T. LK3/L (Zircaloy-2)* wt% 1.34 0.18 0.11 0.05 0.34

• 1320

70 logA*

• 14.2

LK3/L (Zircaloy-2) at% 1.02 0.29 0.19 0.077 0.557 Low-tin Zircaloy-4 wt% 1.20 0.22 0.107

• 0.337

140 1730

• 504°C

SRA Low-tin Zircaloy-4 at% 0.914 0.356 0.186

• 0.542

Zr-2.5%Nb wt%

• 0.07
• 2.5
• 180

1170 60 500°C PRX Zr-2.5%Nb at%

The chemical composition of alloys used for the current study, (H.T.: heat treatment, SRA: stress relieve annealed, and PRX: partially recrystallized condition) [13], [17] and [19]

Materials selected for the current study

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

Step 1

• i. Study a very high burnup cladding to look for great changes (in this case a Zirc-2

with LK3/L grade from KKL after 9 cycles) − a- Observation of macroscopic changes in the cladding from PIE: NDT and DT − b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB, micropillar compression test, etc.

a- Observation of macroscopic changes in the cladding from PIE: NDT and DT

Page 9

500 1000 1500 2000 2500 3000 3500 4000 9,60 9,65 9,70 9,75 9,80 9,85 9,90 9,95 DIM0.180 DIM90.270 DIM45.225 DIM135.315 DIMRING

Filename: AGB108-G6_272-25167 Profilometrie.opj

Diameter [mm] Profilometry A GB108-G6 (272-25167; LK3) Position from tip of bottom end plug [mm]

WI43-23.07.2007

9,5 9,6 9,7 9,8 45 90 135 180 225 270 315 9,5 9,6 9,7 9,8

WI43-17.10.2007

Filename: AGB108-G6_272-25167 Profilometrie.opj

Diameter [mm]

750 1507 1995 3000 [mm from BEP]

Diameter midspan AGB108-G6 (272-25167)

Results of profilometry of rod AGB108-G6 diameter plot at different orientations: 0°, 45°, 90°, 135° (corrected for comparison with other data) Azimuthal plot of rod diameter at different elevations.

Comparison of hydrogen content (DT) at different elevations (a) and rod growth (b) from pool side NDT investigations; for LK3/L cladding grade and as a function

• f number of cycles. As can be observed, the 9 cycle rod shows a rapid rod

growth.

a- Observation of macroscopic changes in the cladding from PIE: NDT and DT

a b Ledergerber, G., et al. “Fuel Performance Beyond Design – Exploring the Limits,” Proceedings of 2010 LWR Fuel Performance/Top Fuel/WRFRM, American Nuclear Society, Orlando, FL, Sept 26–29, 2010, Paper 0044

a

Comparison of hydrogen content (DT) and rod growth (NDT), (a) as a function of

• xide thickness, for LK3/L cladding grade and (b) as a function of burnup for two

different cladding grades

b

a- Observation of macroscopic changes in the cladding from PIE: NDT and DT

a

Comparison of hydrogen content (DT) and rod growth (NDT), (a) as a function of

• xide thickness, for LK3/L cladding grade and (b) as a function of burnup for two

different cladding grades

b

a- Observation of macroscopic changes in the cladding from PIE: NDT and DT

0.00 0.20 0.40 0.60 0.80 1.00 200 400 600 800 Rod growth (%) Hydrogen content in metal (ppm)

Correlation between rod average hydrogen concentration in the metal and %age rod growth

LK3/L rods LK2/L rods LK3/L 3c

Calculated volume increase of the rod as function of hydrogen content (all expansion assumed in axial direction)

Step 1

• i. Study a very high burnup cladding to look for great changes (in this case a Zirc-2

with LK3/L grade from KKL after 9 cycles) − a- Observation of macroscopic changes in the cladding from PIE: NDT and DT − b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB, micropillar compression test, etc.

Fe Ni O Cr

AGB108-G6-GF: SEM, BSE and EPMA distribution mappings of Cladding at the Metal/Oxide interface; at orientation of (2000). HV 15kV, 200nA, 120 ms/pix (192 x 192 pix.) Large hydrides in the metal side, depleted from Fe and to some extent Ni.

b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB…

• A. Baris et al. «Causes of increased corrosion and hydrogen uptake of Zircaloy-2 cladding at high burnups – A comparative

study of the chemical composition of a 3 cycle and a 9 cycle cladding»; Proceedings of the TopFuel Conference, 2018.

Distribution of hydride precipitates, in the 9-cycle LK3/L observed in the EPMA, with no etching. Hydrides are distributed unevenly and the waterside rim has a higher density of hydrides.

b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB, micropillar compression test, …

Area fraction of hydrides along the 9-cycle LK3/L cladding wall from inner to the

• uter surface of the tube. The hydride present in the liner is not taken into

consideration in this graph.

b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB, micropillar compression test, …

HAADF and Chemi-STEM maps of 9-cycle LK3/L showing the grain boundary segregation in the metal side of the interface [19].

b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB …

• A. Baris et al. «Causes of increased corrosion and hydrogen uptake of Zircaloy-2 cladding at high burnups – A comparative

study of the chemical composition of a 3 cycle and a 9 cycle cladding»; Proceedings of the TopFuel Conference, 2018.

3D imaging of the FIB cut layers (example of the metal-

• xide region in the cladding)

Page 18

Cracks of the oxide formed in the last cycle, i.e. close to the metal-oxide interface. A, B and C indicates the possible subdivision of the cracks.

radial crack metal-oxide interface A B C

• A. Baris et al. “Chemical and microstructural characterization of a 9 cycle Zircaloy-2 cladding using EPMA and FIB

tomography”; Journal of Nuclear Materials, 2018, 504, pp. 144-160.

Page 19

3D reconstruction of the waterside oxide of the 9 cycle sample. Green

• bjects: cracks, transparent pink object: oxide matrix.

3D imaging of the FIB cut layers (example of the metal-

• xide region in the cladding)
• A. Baris et al. “Chemical and microstructural characterization of a 9 cycle Zircaloy-2 cladding using EPMA and FIB

tomography”; Journal of Nuclear Materials, 2018, 504, pp. 144-160.

Page 20

The 3D reconstruction of the metal-oxide interface of the 9 cycle material corresponding to the red frame in part C of Figure 4. Green objects: cracks in the

• xide. Hydrides in the metal are shown in different colours.

3D imaging of the FIB cut layers (example of the metal-

• xide region in the cladding)

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

Step 2

• Step-2: Then extrapolate the changes to lower burnup examples of the same

cladding grade under very similar conditions − Will only report the recent destructive results: − Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB, micropillar compression test, etc.

Page 23

3D reconstruction of the oxide of the 3-cycle sample. Green objects: cracks, transparent pink object: oxide matrix.

3D imaging of the FIB cut layers (example of the metal-

• xide region in the cladding)…

Comparison with lower burnup materials

• A. Baris et al. “Chemical and microstructural characterization of a 9 cycle Zircaloy-2 cladding using EPMA and FIB tomography”; Journal of

Nuclear Materials, 2018, 504, pp. 144-160.

Page 24

3D imaging of the FIB cut layers (example of the metal-

• xide region in the cladding)…

Comparison with lower burnup materials

3- c 9- c

Oxide layer Hydride and oxide layer

Microcrack distribution in the oxide (upper tomograms) as a function of number of cycles in the reactor, and the distribution of hydrides in the metal side of the interface (lower tomogram) indicating the large hydride objects.

b- 3D- Distribution of hydrides in the metal and the microcracks in the oxide … Comparison with lower burnup materials

The correlation of volume fraction of hydrides, as a function of volume fraction of micropores in the oxide, at 1 micrometer on the two sides of the metal-oxide interface [30].

b- 3D- Distribution of hydrides in the metal and the microcracks in the oxide … Comparison with lower burnup materials; Quantification of volume fractions

Baris A., PhD thesis., 2019.

(a) grain boundary segregation as a function of number of cycles represented as the ratio of alloying element in the GB with respect to the matrix. Cr segregation was observed only at limited sites of the GB after 9 cycles.(b) Data from Fig. (a) normalized using the matrix concentrations in EPMA. For archive material, the matrix composition from TEM is used. For 7-cycle sample (no EPMA available), the EPMA from 6 cycle material is used. The composition of the “metal at the interface” region was used from the EPMA of the 3 and 6 cycle samples. 9-cycle EPMA data from the “bulk metal” region [30].

b- Observation of microscopic changes in the cladding: from EPMA, TEM, 3D FIB, micropillar compression test, … Comparison with lower burnup materials

Baris A., PhD thesis., 2019.

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

I. Examination of Mechanical properties: Micro-compression results

Example of 9-cycle cladding

Metal Pillar at M/O Oxide Pillar at M/O

[A. W. Colldeweih, A. Baris, P. Spätig, S. Abolhassani; “Evaluation of Mechanical Properties of Irradiated Zirconium Alloys at the Metal-

• xide Interface”; Materials Science & Engineering A, 2019; 742, pp.842-850.

30

Material Metal- E (GPa) Metal- YS (GPa) Metal strain (Y) YS/E Oxygen rich region- E (GPa) Oxygen rich region- YS (GPa) Oxygen rich region- strain (Y) YS/E Oxide- E (GPa) Oxide- YS (GPa) Oxide Strain (Y) YS/E 3-cycle 62.3 1.1 0.018 39 1.3 0.033 26-106 1 - 5.1 0.076 9-cycle 39.5 0.9 0.023 36.5 1.5 0.041 23 - 60 0.1 - 3.2 0.026 Fuel Cladding unirr [Yagnik] 95-103 0.6

• Fuel cladding irr

[Yagnik]

• 0.8
• Oxide bulk [refxx]

190-200 Oxide bulk [refyy] 100-250 0.12-0.78 /(1.2-5)*

I. Examination of Mechanical properties: Micro-compression results

Mechanical properties of micro-pillars at Metal-Oxide interface. LK3/L 3-cycle and 9-cycle claddings. Previous measurements from the literature or the general information about zirconia provided for comparison.

*The values between parentheses show the compressive strength for comparison

[A. W. Colldeweih, A. Baris, P. Spätig, S. Abolhassani; “Evaluation of Mechanical Properties of Irradiated Zirconium Alloys at the Metal-oxide Interface”; Materials Science & Engineering A, 2019; 742, pp.842-850.

Comparison of yield strengths between 3 and 9-cycle LK3/L

Page 31

I. Examination of Mechanical properties: Micro-compression results

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

SEM micrographs from 3, 6 and 9-cycle LK3/L materials at the metal-oxide interface, during the FIB cuts showing the charging effects and the small non-charging layer close to the metal-oxide interface. The thickness of this very small layer seems to be rather similar for the material after the different residence times.

• II. Examination of Semi-conducting properties:

SEM observations

550 nm 590 nm 600 nm

3- c 6- c 9- c Baris A., PhD thesis., 2019.

C- Micromanipulators for conductivity measurement

Page 34

In this example, the metal electrode has been created at an angle

• f 90° w.r.t. the wedge, for the ease of access.

New method to determine electrical properties of the material at the micro and nanometric level

Radovan Vanta et al.: “Examination of semiconducting properties of oxides in the vicinity of metal-oxide interfaces for selected alloys”; The 16th European Microscopy Congress 2016; DOI: 10.1002/9783527808465.EMC2016.6954

a

Examples of resistivity calculations from the measurement of semi-conducting properties of the oxide in the vicinity of interface by means of micromanipulators in the SEM. (a) Results for LK3/L; 3 and 9-cycle claddings (BWR). (b) the same measurements for low-tin Zircaloy-4 and Zr2.5%Nb materials from PWR.

• II. Examination of Semi-conducting properties:

Micromachining and micromanipulator observations

Hawes et al. To be submitted.

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

37

O Fe Cr Sn

An example of EPMA map for low-tin Zircaloy-4 and the distribution of alloying elements in the metal, oxide and along the metal-oxide interface. Higher concentration of Fe in the metal side of the interface is observed. The histograms in the inset indicate the distribution of alloying elements in the different regions of the metal and oxide.

Observation of microscopic changes in the cladding: from EPMA … Comparison with PWR materials

The correlation of volume fraction of hydrides, as a function of volume fraction of micropores in the oxide, at 1 micrometer on the two sides of the metal-oxide interface for all claddings from BWR and PWR [30].

b- 3D- Distribution of hydrides in the metal and the microcracks in the oxide … All materials studied represented; Quantification of volume fractions

Baris A., PhD thesis.

a b

Examples of resistivity calculations from the measurement of semi-conducting properties of the oxide in the vicinity of interface by means of micromanipulators in the SEM. (a) Results for LK3/L; 3 and 9-cycle claddings (BWR). (b) the same measurements for low-tin Zircaloy-4 and Zr2.5%Nb materials from PWR.

• II. Examination of Semi-conducting properties:

Micromachining and micromanipulator observations All materials studied represented; Quantification of resistivity of different oxide regions

Radovan Vanta et al.: “Examination of semiconducting properties of oxides in the vicinity of metal-oxide interfaces for selected alloys”; The 16th European Microscopy Congress 2016; DOI: 10.1002/9783527808465.EMC2016.6954

Outline

• i. Study of the 9-cycle Zircaloy-2, LK3/L from KKL (BWR) cladding to look for great

changes

• ii. Comparison with a low burnup (here 3-cycle) and an intermediate burnup

(here 6-cycle) cladding from the same grade material and the same reactor; to verify the validity of indicators

• iii. Examination of mechanical and semi-conducting properties of the oxide layers

in the case of the two extreme cases

• iv. Verification of PWR material to search for general phenomena
• v. The use of the data to define the determinant indicators
• vi. Correlation with the known parameters
• vii. Conclusion

• A. microscopic factors

− i- the chemical composition of the metal and the oxide − ii- the distribution of hydrides near the metal oxide interface − iii- the volume fraction of microcracks in the oxide side of the interface − iv- the extent of grain boundary segregation close to the metal-oxide interface − v- the coupled mechanical properties of the interface − vi- the semi-conducting properties of the oxide sublayer and its extent

• B. macroscopic factors

− vii- pellet clad interactions − viii- dimensional changes and increase of rod radius − ix- rod growth − x- inner clad-wall compression through constant pellet swelling − xi- pellet swelling and fission gas release

Determination of Indicators

41

Metal Pillar at M/O

a b c d d e

a- Spongy oxide, b- Dense oxide, c- Inter-phase layer, d- Oxygen rich layer, e- Metal sublayer

• Suggest at least three stages:
• Stage 1: A very tenacious oxide, the oxidation governed by diffusion , the

mechanical properties of oxide allow for stresses induced from oxidation and from the cladding dimensional changes, hydrides are small precipitates, extended semi-conducting oxide at the metal-oxide interface (e.g. 3-cycle LK3/L)

• Stage 2: Oxide composition has changed, the oxide has developed microcracks,

the oxidizing species percolate into the inner layer, the gradual dimensional changes of the cladding increase due to growth from different origins (e.g. irradiation and hydride precipitation), the mechanical properties of oxide change and the oxide is less resistant to crack formation, hydride rims induce more microcracks, the semi-conducting properties will evolve.

• Stage 3: The indicators reach their thresholds, and the above mentioned

behaviors become macroscopically measureable.

Stages of oxidation and hydrogen uptake in-reactor

Page 43

Page 44

Wir schaffen Wissen – heute für morgen

Our thanks go to

• PSI
• KKL
• KKG
• Swissnuclear
• Westinghouse
• Areva
• MUZIC-3

Power History example of AGB108-G6

Page 45

Typical power history of the 9 cycle LK3/L cladding showing six different elevations (here called nodes). The local burnup is provided in the inset for each elevation [14].

Ledergerber, G., et al. “Fuel Performance Beyond Design – Exploring the Limits,” Proceedings of 2010 LWR Fuel Performance/Top Fuel/WRFRM, American Nuclear Society, Orlando, FL, Sept 26–29, 2010, Paper 0044

The IFA-743 report is EPRI report number 3002015313

Page 46

Heat of solution of Hydrogen in transition metals

Page 47

• A. Andreasen, RISO, Report, Denmark

EPRI research

Page 48

Christensen et al. 2017, restricted use.

49

Brankov, V., et al. Nuclear Engineering and Design; vol. 305 (2016) 559–568. (RFTHF): ratio of fast to thermal neutron flux