- Email: [email protected]
Melting test of penetrating tube through BWR-RPV bottom wall
Annals of Nuclear Energy 118 (2018) 212–219
Contents lists available at ScienceDirect
Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene
Melting test of penetrating tube through BWR-RPV bottom wall Masanori Naitoh a,⇑, Hiroyuki Suzuki a, Marco Pellegrini a, Sang Mo An b, Jin Ho Song b a b
The Institute of Applied Energy, 1-14-2, Nishi-shinbashi, Minato-ku, Tokyo 105-0003, Japan Korea Atomic Energy Research Institute, Daedeokdaero, Yuseong, Daejeon 305-353, Republic of Korea
a r t i c l e
i n f o
Article history: Received 30 November 2017 Received in revised form 22 March 2018 Accepted 23 March 2018 Available online 16 April 2018 Keywords: Severe accident In-core monitor guide tube (ICMGT) Fukushima Daiichi NPP Melt behavior
a b s t r a c t Melt behavior of a BWR in-core monitor (ICM) guide tube, which penetrates a RPV bottom wall, under a severe accident condition was tested using a corium with a mixture of UO2, ZrO2 and Zr. There are several BWR-ICMs; two of them are a short range monitor (SRM) and an intermediate range monitor (IRM), which are made of stainless steel, have the same tube dimensions and the bottom ends of the tubes open to the drywell (or, the pedestal region). In the test, the full scale SRM/IRM guide tube with the same dimension as one of Fukushima Daiichi Unit-1 but short length, was installed in the lower crucible which simulated the lower plenum and which bottom simulated the RPV bottom wall with about 200 mm thickness. The melt corium was generated in the upper crucible and it was discharged into the lower crucible. The test result showed that: (1) The corium discharged into the lower plenum attacked the guide tube first resulting in the tube melt from the top and then accumulated in the lower plenum. (2) Small amount of corium had fallen out from the bottom opening of the tube into the pedestal as particles, and such outflow stopped soon because a stainless steel melt had solidified in the tube and blocked the flow path. (3) The depth of the melt ingression into the tube was 60–65 mm from the inner surface of the RPV bottom wall, where the tube was completely blocked by the solidification of steel melt and no oxide melts had ingressed there. Ó 2018 Elsevier Ltd. All rights reserved.
1. Introduction The Fukushima Daiichi Nuclear Power Plant Unit-1, -2, and -3 were severely damaged after the huge Tsunami followed by the earth quake occurred at March 11, 2011. Core meltdowns occurred at all the units. The radioscopic images of the Unit-1 reactor core were taken by the cosmic muon radiography with the attenuation method Takasaki et al. (2015)1. The images showed that the Unit-1 reactor core was almost empty, which suggested that almost all of the fuel debris had fallen down from the reactor pressure vessel (RPV) onto the pedestal floor.1 Recent investigation inside the primary containment vessel (PCV) of Unit-1 also suggested the existence of fuel debris inside the pedestal.2 Similar investigations inside the
PCVs of Unit-2 and Unit-3 showed piles of fuel debris on the pedestal floors.3,4 The question arose: how a large amount of corium had fallen down. The question still remains. That is, from which a large amount of corium had fallen down; from the break location due to melt, or creep of the RPV bottom wall itself, or, due to melt of the in-core monitor guide tubes which penetrate the RPV bottom wall. There are many tubes which penetrate the BWR lower head as shown in Fig. 1; control rod guide tubes (CRGTs) and in-core monitor guide tubes (ICM-GTs). Two of the ICMs are a short range monitor (SRM) and an intermediate range monitor (IRM): the configurations of their guide tubes are the same and the bottom ends of the tubes open to the pedestal region. The other ICM-GTs than SRM/IRM guide tubes are bottom closed. When high temperature corium accumulates in the lower plenum under a severe accident condition,
⇑ Corresponding author. E-mail address: [email protected] (M. Naitoh). Open information obtained from the web site; http://photo.tepco.co.jp/en/date/ 2015/201503-e/150319-01j.html, updated at March 19,2015. 2 Open information obtained from the web site; http://www.tepco.co.jp/en/ nu/fukushima-np/handouts/2017/images/handouts_170327_01-e.pdf, updated at March 27, 2017. 1
https://doi.org/10.1016/j.anucene.2018.03.034 0306-4549/Ó 2018 Elsevier Ltd. All rights reserved.
3 Open information obtained from the web site; http://www.tepco.co.jp/en/ nu/fukushima-np/handouts/2018/images/handouts_180119_01-e.pdf, updated at January 19, 2018. 4 Open information obtained from the web site; http://www.tepco.co.jp/en/ nu/fukushima-np/handouts/2017/images/handouts_171130_03-e.pdf, updated at November 30. 2017.
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
Fig. 1. Pressure boundary.
it may melt the SRM/IRM guide tubes made of stainless steel, resulting in corium release into the pedestal, before the melting of the RPV bottom wall. The test has been implemented in order to examine melt behavior of the SRM/IRM guide tube and behavior of melt ingression into the tube. In the test, a full sized SRM/IRM guide tube with the same dimension as one of Fukushima Daiichi Unit-1 but short length was used and a melt with a mixture of UO2, ZrO2 and Zr was used as a corium. This paper describes the test results. 2. Possible accident progression at the Fukushima Daiichi Nuclear Power Plant The decay heat was removed by the cooling systems at early phase after the reactor scrams at Units 1, 2, and 3. The reactor pres-
213
sures rose to about 7 MPa and were kept almost constant by steam release through the safety relief valves after the stop of the decay heat removal systems, resulting in water level decrease in the reactor cores. After the core uncovery, fuel temperatures rose due to decay heat generation and core meltdowns occurred. Some of melt materials, corium, might flow down continuously into the lower plenums of the RPVs from the core regions, but many of them was considered to be retained on the lower core plate for a while. Then, it was supposed that the corium on the lower core plate slumped down into the lower plenum. There were a lot of penetration tubes in the lower plenum as typically shown in Fig. 1. The top ends of them were closed. The bottom ends of them were also closed except the SRM/IRM guide tubes of which bottom ends were open to the pedestal. Supposing formation of corium pool with the temperature of around 2700 K in the lower plenum, the SRM/IRM guide tubes made of stainless steel with the melting temperature of around 1750 K were considered to be damaged earlier than the other tubes, since their tube wall thickness was the most thinnest. There were ten-odd SRM/ IRM guide tubes Installed distributedly at every Unit 1, 2, and 3. The SRM/IRM guide tubes might be damaged successively, or when once one tube was first damaged, all the corium in the lower plenum might fall down through the damaged portion, depending on amount of corium pool in the lower plenum. This paper focuses on the behavior of the single SRM/IRM guide tube melt and ingression of melt materials into the tube, supposing the corium formation in the lower plenum. Some of the SRM/IRM guide tubes penetrated the inclined RPV bottom heads, but only a single tube through the flat plate was used as the test specimen.
3. Test facility 3.1. Test specimen and test vessel Fig. 2 shows the test specimen of the SRM/IRM guide tube. The dimensions of the cross-section of the test specimen is the same as ones of Fukushima Daiichi Unit-1, but its length was short. The SRM/IRM guide tube was inserted in the core region in the actual plant, but the length of the test specimen was limited to 336 mm. Similarly, the length of the actual tube below the RPV bottom
Fig. 2. Test specimen of SRM/IRM guide tube.
214
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
cible simulated the pedestal region. Pressure and temperature in each vessel were measured. 3.2. Corium composition At Fukushima Daiichi accident, the metallic components, mainly stainless steel, with a lower melting temperature in the core region should melt first and relocate to the lower plenum, then oxidic components, UO2 and ZrO2, with higher melting temperature should relocate. Since the melting temperatures of the ICM-GTs and the RPV bottom wall itself were almost the same as one of the metallic components in the core region, they would not melt during the first relocation process of the melts of the metallic components in the core region, considering heat loss outside the RPV. The situation of relocation of the oxidic component melts was simulated in the test. The initial amounts of UO2 and Zr in the Unit-1 core before the accident were about 69 103 kg and 26.2 103 kg respectively. Supposing all Zr was oxidized under the severe accident condition, the amount of ZrO2 became 30.8 103 kg. Therefore, the ratio of amounts of UO2 and ZrO2 was nearly 1:0.44 under the severe accident condition. 3.3. Test procedures 3.3.1. Corium generation Corium (melt) was generated in the upper crucible. Initially UO2 and ZrO2 pellets were used as the material including Zr ring as an initiator for melt spread, which was put on the center of the crucible. The weight of each composition was:
UO2 : ZrO2 : Zr ¼ 105 : 45 : 0:3 ðkgÞ The materials in the upper crucible were indirectly heated by a high frequency induction heating system. When the enough amount of corium was generated, the induction heating was switched off and the melts were discharged into the lower crucible.
Fig. 3. Thermocouple location.
wall was about 4 m, but the test specimen had only 310 mm. The material of the test specimen was the same as one of the actual plant, that is a stainless-steel. In the actual guide tube, a neutron sensor and its lead wires were inserted inside, but inside the test specimen was empty except thermocouples and their support clamps. The tube support flange and the support for thermocouple (T/C) lead wires were fixtures specific to the test specimen. The tube penetrated at the center of the test RPV bottom wall, with the weld portion only at inside. The reactor water flows in the annulus flow path between the inner and the outer tubes, which bottom end was blocked, under a normal operating condition of the plant, but the annulus flow path in the test specimen was open at the bottom to see melt ingression behavior through the annulus flow path. Fig. 3 shows the thermocouple (T/C) locations. Temperatures of the tube were measured at the 7 axial locations. Two T/Cs in opposite radial directions were installed at each axial location. Six T/Cs were axially inserted in the RPV bottom wall. The test was conducted at Korea Atomic Energy Research Institute (KAERI). Fig. 4 shows the test vessel, called VESTA. There was the upper crucible in the upper vessel, where the melt was generated by induction heating. The lower crucible in the lower vessel simulated the lower plenum inside the RPV. The test specimen was installed in the lower crucible. The vessel below the lower cru-
3.3.2. Discharge of corium into the lower crucible and its heating The bottom of the upper crucible was punched to discharge the corium into the lower crucible where the test specimen was installed. In order to keep the melt condition of the corium in the lower crucible, the induction heating was switched from the upper crucible to the lower crucible. 4. Result Applied power to the upper crucible for corium generation is shown in Figs. 5 and 6 shows the temperature transient on the surface of the upper crucible measured by a pyrometer. The quality factor (Q-factor) shown in Fig. 5 is an index to judge the melting situation. The detailed definition of the Q-factor is written in the Reference Hong et al. (2003). The melting state can be predicted from the change of the Q-factor, since the electrical conductivity of the oxides increases with temperature. The lower limit of temperature measurement by the pyrometer was 1500C. The temperature fluctuated several times during the transient. This was due to relocation of un-melted low temperature pellets directly under the pyrometer. Since there were some aerosols observed in the upper crucible, the temperature indication by the pyrometer might be a little lower than the actual temperature. Actually it is said that the liquidus temperature of the mixture of UO2 + ZrO2 is about 2800 K (2527C). The KAERI operators of the test facility judged that a sufficient amount of melt was generated at about 4200 s from the change of the Q-factor, based on their experiences so far. The bottom of the upper crucible was punched
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
215
Fig. 4. Test vessel (VESTA).
to discharge the melt to the lower crucible at 4213 s, and at the same time the application of power into the upper crucible was stopped. The total discharged amount of corium was 70 kg. Some un-melted pellets remained in the upper crucible as shown in Fig. 7. Fig. 8 shows the applied power to the lower crucible. The heating of the lower crucible was delayed due to un-expected trouble
and it started at 4340 s, 127 s after the stop of heating to the upper crucible. The heating was interrupted for 16 s after continuous heating for 510 s due to electrical short circuit. Then the heating continued for 774 s until the second interruption for 22 s. Such delay and interruption of heating did not affect the melt behavior of the guide tube in the lower crucible since it melted earlier before the time of 4380 s, as described below in Fig. 10. The decay heat
216
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
Fig. 5. Applied power to the upper crucible. Fig. 8. Applied power to the lower crucible.
Fig. 6. Temperature measurement by pyrometer.
Fig. 7. Upper crucible after melt discharge.
Fig. 9. Temperature and pressure transients in vessels.
generation from the high temperature fuel corium continues at the actual plant, but such long term transient could not be tested since similar electrical short circuits occurred several times, and finally the heating was completely stopped at 6708 s in this test. Fig. 9 shows the temperature and pressure transients in the vessels. These transients were very specific to the test facility. During the high temperature melt discharge process from the upper crucible, it temporarily spread on the intermediate melt catcher (see Fig. 4. Since the spread area was wide, Argon gas in the upper vessel was temporally heated up to 539C mainly due to radiant heat transfer. Due to this heating, the upper vessel pressure also increased to 0.212 MPa. After melt discharge to the lower crucible, the temperature and pressure in the lower vessel also temporally increased up to about 160C and 0.12 MPa. Since the melt discharge from the bottom open end of the guide tube to the pedestal was very small, the temperature and pressure change in the pedestal were also small.
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
217
Fig. 10. Temperature distribution of the tube inside the RPV.
Fig. 13. Falling down of particle corium from inner tube.
Fig. 11. Falling down of particle corium from annulus.
Fig. 12. Crust accumulated on the support for the T/C wires.
Fig. 10 shows the temperature distribution of the tube inside RPV. The time of melt discharge into the lower crucible was 4213 s. The upper limit of the temperature measurement by thermocouples was 1500C which was comparable to the melt temperature of the steel. The temperature measurement was largely fluctuated after it reached to about 1500C. This was due to the melt of the outer tube at this timing. The outer tube had initially melted at the position of KP-16 at 4327 s. Next, the position of KP-15 with the same elevation as KP-16 had melted at 4332 s, 5 s after the melt of the position of KP-16. This was due to the direct impingement of particulate corium to the top of the tube when the corium was discharged from the upper crucible. Then the locations of KP17 and 18 melted at 4335 s, finally KP-19 and 20 at 4371 s. The corium had first started to flow down from the bottom open end of the annulus flow path as particles at almost the same timing of KP-16 melt, 4327 s, followed by flowing down from the bottom open end of the inner tube at 4335 s. Fig. 11 shows the picture of particle corium falling down from the annulus. The duration of such falling down was very short, only several seconds. Fig. 12 shows the picture taken after the melt test. Small amount of crust accumulated on the support for T/C lead wires. They looked like a mixture of oxides and steel.
218
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
Fig. 14. Temperature distribution of the tube below the RPV wall.
Fig. 15. Temperature distribution in the RPV wall.
Fig. 16. Particle debris remained in the tube. Fig. 17. Melt ingression into he tube.
Fig. 13 shows the picture of particle corium falling down from the inner tube. It looked like a continuous flow, but actually they were particles. The duration of such falling down was also very short, only the order of ten seconds.
Fig. 14 shows the temperature distribution of the tube below the RPV wall. It should be first noted that the maximum temperature, which appeared at KP-21, 65 mm below the RPV inner sur-
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
face, was less than 1000C and that the tube did not melt. Second, it should be noted that the temperatures showed large deviation even at the same elevation. The temperatures at KP-21 and KP23 were higher than ones at KP-22 and KP-24. The reverse trend was shown below KP-23 and KP-24. Based on this temperature behavior, it was estimated that the corium flow did not meet the flow path in the tube, but it flowed down as particles. Fig. 15 shows the temperature distribution in the RPV wall. Since the RPV wall received heat from the corium on the wall and at the same time released heat radially and axially, the temperature was higher at closer location to the inner surface of the RPV wall. The measured maximum temperature was 581C, which was well below the melting temperature of the steel. 5. Post test inspection Corium ingression into the tube was inspected. As stated, the corium was considered to flow down in the tube as particles. Some of them were caught on the T/C support clamps as shown in Fig. 16. The pictures were taken by a borescope which was inserted from the bottom open end of the pipe. They looked like a metal (steel). Nothing had deposited on the tube wall below the RPV wall. The tube inside the RPV had perfectly melted. The melt material solidified on the surface of the RPV and was hardly removed from the RPV wall. The test tube was welded at the inner surface of the RPV wall as shown in Fig. 2, and the test tube was anchored to the RPV at the weld portion. The weld portion was covered by the solidified material and was integrated with it, resulting in one large ingot. Some melt flowed into the tube and solidified, resulting in blockage of the flow path in the tube. The depth of the melt ingression from the inner surface of the RPV wall was 60 mm at the annulus region and 65 mm at the inner tube as shown in Fig. 17. Several samples were taken from the solidified materials; solidified material on the RPV inner surface, bottom of the ingot in the inner tube, and released materials from the annulus and dropped on the support for the T/C wires. The samples were analyzed by using ICP-AES (Inductively Coupled Plasma-Atomic Emission Spectrometry). The chemical analysis result showed very small amount of uranium, less than 0.016 wt% (0.16 mg/g), and ferrous component of more than 70 wt% for all the samples. It was concluded from the chemical analysis that the stainless steel tube melted in the lower plenum and some of the melted steel dropped down through the bottom of the tube and the material which irrupted into the tube soon solidified and blocked the tube preventing further corium ingression. 6. Conclusion A melt behavior of a BWR in-core monitor guide tube (ICM-GT) for a short range monitor (SRM) and an intermediate range monitor (IRM), which penetrates a RPV bottom wall, under a severe accident condition was tested using a corium with a mixture of UO2, ZrO2 and Zr. The test specimen was a full scale SRM/IRM guide tube with the same dimension as one of Fukushima Daiichi Unit-1 but short length. The melt corium was generated in the upper crucible and it was discharged into the lower crucible where the test specimen was installed. The test result showed that:
219
(1) When the corium was discharged into the lower crucible, the guide tube inside the RPV was first attacked by the corium and then it accumulated in the lower crucible. (2) Due to the first attack of corium, the guide tube in the lower crucible melted from the top. (3) Some of the melts were caught on the T/C support clamps as particulates. (4) Small amount of melts had fallen out from the bottom open end of the tube into the pedestal as particles, and such outflow stopped soon because a stainless steel melt had solidified in the tube and blocked the flow path. (5) The depth of the melt ingression into the guide tube was 60– 65 mm from the inner surface of the RPV bottom wall, where the tube was completely blocked by the solidification of steel melt and no oxide melts had ingressed there. In the test, the time of induction heating to the lower crucible was limited to only about 23 min. At the actual plant, on the other hand, the high temperature corium with decay heat generation might stay longer in the lower plenum resulting in heating the RPV bottom wall. In such a case at the actual plant, the depth of the melt (steel melt) ingression into the tube might become much longer, but the possibility of ingression of oxide melts was considered to be small because the tube might be blocked by the solidification of steel melt due to heat radiation from the tube outside the RPV. Moreover, the guide tube ejection from the RPV wall might occur due to melting of the weld portion at the inner surface of the RPV by heat conduction from the high temperature corium. 7. Future work This paper describes only the test result. The important result obtained from the test was that the tube was fully blocked by the ingression of the melt of stainless steel guide tube and its solidification. At the actual plant, detailed damage behavior of the guide tube can be evaluated by heat conduction calculation from the high temperature fuel corium with decay heat generation. The accident progression analysis of the actual plant shall be conducted considering such heat conduction calculation. Acknowledgements The test has been jointly conducted by the Institute of Applied Energy and Korea Atomic Energy Research Institute with a financial sponsorship of the Japanese Ministry of Economy, Trade and Industry. The test specimen was provided by Hitachi-GE Nuclear Energy, Ltd. The authors would like to thank Dr. Jaebong Lee of KAERI for the valuable discussions. The authors would also like to thank test engineers of KAERI for the implementation of the tests. References Hong, S.W. et al., 2003. Application of cold crucible for melting of UO2/ZrO2 mixture. Mater. Sci. Eng., 297–303 Takasaki, Fumihiko, Mizokami, Shinya, Nagano, Mamoru, 2015. Inspection of the Damaged Reactor at the Fukushima Daiichi with the Cosmic Muon Radiography Muon Investigation Report, 2015 July (In Japanese). IRID (International Research Institute for Nuclear Decommissioning).
Contents lists available at ScienceDirect
Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene
Melting test of penetrating tube through BWR-RPV bottom wall Masanori Naitoh a,⇑, Hiroyuki Suzuki a, Marco Pellegrini a, Sang Mo An b, Jin Ho Song b a b
The Institute of Applied Energy, 1-14-2, Nishi-shinbashi, Minato-ku, Tokyo 105-0003, Japan Korea Atomic Energy Research Institute, Daedeokdaero, Yuseong, Daejeon 305-353, Republic of Korea
a r t i c l e
i n f o
Article history: Received 30 November 2017 Received in revised form 22 March 2018 Accepted 23 March 2018 Available online 16 April 2018 Keywords: Severe accident In-core monitor guide tube (ICMGT) Fukushima Daiichi NPP Melt behavior
a b s t r a c t Melt behavior of a BWR in-core monitor (ICM) guide tube, which penetrates a RPV bottom wall, under a severe accident condition was tested using a corium with a mixture of UO2, ZrO2 and Zr. There are several BWR-ICMs; two of them are a short range monitor (SRM) and an intermediate range monitor (IRM), which are made of stainless steel, have the same tube dimensions and the bottom ends of the tubes open to the drywell (or, the pedestal region). In the test, the full scale SRM/IRM guide tube with the same dimension as one of Fukushima Daiichi Unit-1 but short length, was installed in the lower crucible which simulated the lower plenum and which bottom simulated the RPV bottom wall with about 200 mm thickness. The melt corium was generated in the upper crucible and it was discharged into the lower crucible. The test result showed that: (1) The corium discharged into the lower plenum attacked the guide tube first resulting in the tube melt from the top and then accumulated in the lower plenum. (2) Small amount of corium had fallen out from the bottom opening of the tube into the pedestal as particles, and such outflow stopped soon because a stainless steel melt had solidified in the tube and blocked the flow path. (3) The depth of the melt ingression into the tube was 60–65 mm from the inner surface of the RPV bottom wall, where the tube was completely blocked by the solidification of steel melt and no oxide melts had ingressed there. Ó 2018 Elsevier Ltd. All rights reserved.
1. Introduction The Fukushima Daiichi Nuclear Power Plant Unit-1, -2, and -3 were severely damaged after the huge Tsunami followed by the earth quake occurred at March 11, 2011. Core meltdowns occurred at all the units. The radioscopic images of the Unit-1 reactor core were taken by the cosmic muon radiography with the attenuation method Takasaki et al. (2015)1. The images showed that the Unit-1 reactor core was almost empty, which suggested that almost all of the fuel debris had fallen down from the reactor pressure vessel (RPV) onto the pedestal floor.1 Recent investigation inside the primary containment vessel (PCV) of Unit-1 also suggested the existence of fuel debris inside the pedestal.2 Similar investigations inside the
PCVs of Unit-2 and Unit-3 showed piles of fuel debris on the pedestal floors.3,4 The question arose: how a large amount of corium had fallen down. The question still remains. That is, from which a large amount of corium had fallen down; from the break location due to melt, or creep of the RPV bottom wall itself, or, due to melt of the in-core monitor guide tubes which penetrate the RPV bottom wall. There are many tubes which penetrate the BWR lower head as shown in Fig. 1; control rod guide tubes (CRGTs) and in-core monitor guide tubes (ICM-GTs). Two of the ICMs are a short range monitor (SRM) and an intermediate range monitor (IRM): the configurations of their guide tubes are the same and the bottom ends of the tubes open to the pedestal region. The other ICM-GTs than SRM/IRM guide tubes are bottom closed. When high temperature corium accumulates in the lower plenum under a severe accident condition,
⇑ Corresponding author. E-mail address: [email protected] (M. Naitoh). Open information obtained from the web site; http://photo.tepco.co.jp/en/date/ 2015/201503-e/150319-01j.html, updated at March 19,2015. 2 Open information obtained from the web site; http://www.tepco.co.jp/en/ nu/fukushima-np/handouts/2017/images/handouts_170327_01-e.pdf, updated at March 27, 2017. 1
https://doi.org/10.1016/j.anucene.2018.03.034 0306-4549/Ó 2018 Elsevier Ltd. All rights reserved.
3 Open information obtained from the web site; http://www.tepco.co.jp/en/ nu/fukushima-np/handouts/2018/images/handouts_180119_01-e.pdf, updated at January 19, 2018. 4 Open information obtained from the web site; http://www.tepco.co.jp/en/ nu/fukushima-np/handouts/2017/images/handouts_171130_03-e.pdf, updated at November 30. 2017.
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
Fig. 1. Pressure boundary.
it may melt the SRM/IRM guide tubes made of stainless steel, resulting in corium release into the pedestal, before the melting of the RPV bottom wall. The test has been implemented in order to examine melt behavior of the SRM/IRM guide tube and behavior of melt ingression into the tube. In the test, a full sized SRM/IRM guide tube with the same dimension as one of Fukushima Daiichi Unit-1 but short length was used and a melt with a mixture of UO2, ZrO2 and Zr was used as a corium. This paper describes the test results. 2. Possible accident progression at the Fukushima Daiichi Nuclear Power Plant The decay heat was removed by the cooling systems at early phase after the reactor scrams at Units 1, 2, and 3. The reactor pres-
213
sures rose to about 7 MPa and were kept almost constant by steam release through the safety relief valves after the stop of the decay heat removal systems, resulting in water level decrease in the reactor cores. After the core uncovery, fuel temperatures rose due to decay heat generation and core meltdowns occurred. Some of melt materials, corium, might flow down continuously into the lower plenums of the RPVs from the core regions, but many of them was considered to be retained on the lower core plate for a while. Then, it was supposed that the corium on the lower core plate slumped down into the lower plenum. There were a lot of penetration tubes in the lower plenum as typically shown in Fig. 1. The top ends of them were closed. The bottom ends of them were also closed except the SRM/IRM guide tubes of which bottom ends were open to the pedestal. Supposing formation of corium pool with the temperature of around 2700 K in the lower plenum, the SRM/IRM guide tubes made of stainless steel with the melting temperature of around 1750 K were considered to be damaged earlier than the other tubes, since their tube wall thickness was the most thinnest. There were ten-odd SRM/ IRM guide tubes Installed distributedly at every Unit 1, 2, and 3. The SRM/IRM guide tubes might be damaged successively, or when once one tube was first damaged, all the corium in the lower plenum might fall down through the damaged portion, depending on amount of corium pool in the lower plenum. This paper focuses on the behavior of the single SRM/IRM guide tube melt and ingression of melt materials into the tube, supposing the corium formation in the lower plenum. Some of the SRM/IRM guide tubes penetrated the inclined RPV bottom heads, but only a single tube through the flat plate was used as the test specimen.
3. Test facility 3.1. Test specimen and test vessel Fig. 2 shows the test specimen of the SRM/IRM guide tube. The dimensions of the cross-section of the test specimen is the same as ones of Fukushima Daiichi Unit-1, but its length was short. The SRM/IRM guide tube was inserted in the core region in the actual plant, but the length of the test specimen was limited to 336 mm. Similarly, the length of the actual tube below the RPV bottom
Fig. 2. Test specimen of SRM/IRM guide tube.
214
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
cible simulated the pedestal region. Pressure and temperature in each vessel were measured. 3.2. Corium composition At Fukushima Daiichi accident, the metallic components, mainly stainless steel, with a lower melting temperature in the core region should melt first and relocate to the lower plenum, then oxidic components, UO2 and ZrO2, with higher melting temperature should relocate. Since the melting temperatures of the ICM-GTs and the RPV bottom wall itself were almost the same as one of the metallic components in the core region, they would not melt during the first relocation process of the melts of the metallic components in the core region, considering heat loss outside the RPV. The situation of relocation of the oxidic component melts was simulated in the test. The initial amounts of UO2 and Zr in the Unit-1 core before the accident were about 69 103 kg and 26.2 103 kg respectively. Supposing all Zr was oxidized under the severe accident condition, the amount of ZrO2 became 30.8 103 kg. Therefore, the ratio of amounts of UO2 and ZrO2 was nearly 1:0.44 under the severe accident condition. 3.3. Test procedures 3.3.1. Corium generation Corium (melt) was generated in the upper crucible. Initially UO2 and ZrO2 pellets were used as the material including Zr ring as an initiator for melt spread, which was put on the center of the crucible. The weight of each composition was:
UO2 : ZrO2 : Zr ¼ 105 : 45 : 0:3 ðkgÞ The materials in the upper crucible were indirectly heated by a high frequency induction heating system. When the enough amount of corium was generated, the induction heating was switched off and the melts were discharged into the lower crucible.
Fig. 3. Thermocouple location.
wall was about 4 m, but the test specimen had only 310 mm. The material of the test specimen was the same as one of the actual plant, that is a stainless-steel. In the actual guide tube, a neutron sensor and its lead wires were inserted inside, but inside the test specimen was empty except thermocouples and their support clamps. The tube support flange and the support for thermocouple (T/C) lead wires were fixtures specific to the test specimen. The tube penetrated at the center of the test RPV bottom wall, with the weld portion only at inside. The reactor water flows in the annulus flow path between the inner and the outer tubes, which bottom end was blocked, under a normal operating condition of the plant, but the annulus flow path in the test specimen was open at the bottom to see melt ingression behavior through the annulus flow path. Fig. 3 shows the thermocouple (T/C) locations. Temperatures of the tube were measured at the 7 axial locations. Two T/Cs in opposite radial directions were installed at each axial location. Six T/Cs were axially inserted in the RPV bottom wall. The test was conducted at Korea Atomic Energy Research Institute (KAERI). Fig. 4 shows the test vessel, called VESTA. There was the upper crucible in the upper vessel, where the melt was generated by induction heating. The lower crucible in the lower vessel simulated the lower plenum inside the RPV. The test specimen was installed in the lower crucible. The vessel below the lower cru-
3.3.2. Discharge of corium into the lower crucible and its heating The bottom of the upper crucible was punched to discharge the corium into the lower crucible where the test specimen was installed. In order to keep the melt condition of the corium in the lower crucible, the induction heating was switched from the upper crucible to the lower crucible. 4. Result Applied power to the upper crucible for corium generation is shown in Figs. 5 and 6 shows the temperature transient on the surface of the upper crucible measured by a pyrometer. The quality factor (Q-factor) shown in Fig. 5 is an index to judge the melting situation. The detailed definition of the Q-factor is written in the Reference Hong et al. (2003). The melting state can be predicted from the change of the Q-factor, since the electrical conductivity of the oxides increases with temperature. The lower limit of temperature measurement by the pyrometer was 1500C. The temperature fluctuated several times during the transient. This was due to relocation of un-melted low temperature pellets directly under the pyrometer. Since there were some aerosols observed in the upper crucible, the temperature indication by the pyrometer might be a little lower than the actual temperature. Actually it is said that the liquidus temperature of the mixture of UO2 + ZrO2 is about 2800 K (2527C). The KAERI operators of the test facility judged that a sufficient amount of melt was generated at about 4200 s from the change of the Q-factor, based on their experiences so far. The bottom of the upper crucible was punched
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
215
Fig. 4. Test vessel (VESTA).
to discharge the melt to the lower crucible at 4213 s, and at the same time the application of power into the upper crucible was stopped. The total discharged amount of corium was 70 kg. Some un-melted pellets remained in the upper crucible as shown in Fig. 7. Fig. 8 shows the applied power to the lower crucible. The heating of the lower crucible was delayed due to un-expected trouble
and it started at 4340 s, 127 s after the stop of heating to the upper crucible. The heating was interrupted for 16 s after continuous heating for 510 s due to electrical short circuit. Then the heating continued for 774 s until the second interruption for 22 s. Such delay and interruption of heating did not affect the melt behavior of the guide tube in the lower crucible since it melted earlier before the time of 4380 s, as described below in Fig. 10. The decay heat
216
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
Fig. 5. Applied power to the upper crucible. Fig. 8. Applied power to the lower crucible.
Fig. 6. Temperature measurement by pyrometer.
Fig. 7. Upper crucible after melt discharge.
Fig. 9. Temperature and pressure transients in vessels.
generation from the high temperature fuel corium continues at the actual plant, but such long term transient could not be tested since similar electrical short circuits occurred several times, and finally the heating was completely stopped at 6708 s in this test. Fig. 9 shows the temperature and pressure transients in the vessels. These transients were very specific to the test facility. During the high temperature melt discharge process from the upper crucible, it temporarily spread on the intermediate melt catcher (see Fig. 4. Since the spread area was wide, Argon gas in the upper vessel was temporally heated up to 539C mainly due to radiant heat transfer. Due to this heating, the upper vessel pressure also increased to 0.212 MPa. After melt discharge to the lower crucible, the temperature and pressure in the lower vessel also temporally increased up to about 160C and 0.12 MPa. Since the melt discharge from the bottom open end of the guide tube to the pedestal was very small, the temperature and pressure change in the pedestal were also small.
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
217
Fig. 10. Temperature distribution of the tube inside the RPV.
Fig. 13. Falling down of particle corium from inner tube.
Fig. 11. Falling down of particle corium from annulus.
Fig. 12. Crust accumulated on the support for the T/C wires.
Fig. 10 shows the temperature distribution of the tube inside RPV. The time of melt discharge into the lower crucible was 4213 s. The upper limit of the temperature measurement by thermocouples was 1500C which was comparable to the melt temperature of the steel. The temperature measurement was largely fluctuated after it reached to about 1500C. This was due to the melt of the outer tube at this timing. The outer tube had initially melted at the position of KP-16 at 4327 s. Next, the position of KP-15 with the same elevation as KP-16 had melted at 4332 s, 5 s after the melt of the position of KP-16. This was due to the direct impingement of particulate corium to the top of the tube when the corium was discharged from the upper crucible. Then the locations of KP17 and 18 melted at 4335 s, finally KP-19 and 20 at 4371 s. The corium had first started to flow down from the bottom open end of the annulus flow path as particles at almost the same timing of KP-16 melt, 4327 s, followed by flowing down from the bottom open end of the inner tube at 4335 s. Fig. 11 shows the picture of particle corium falling down from the annulus. The duration of such falling down was very short, only several seconds. Fig. 12 shows the picture taken after the melt test. Small amount of crust accumulated on the support for T/C lead wires. They looked like a mixture of oxides and steel.
218
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
Fig. 14. Temperature distribution of the tube below the RPV wall.
Fig. 15. Temperature distribution in the RPV wall.
Fig. 16. Particle debris remained in the tube. Fig. 17. Melt ingression into he tube.
Fig. 13 shows the picture of particle corium falling down from the inner tube. It looked like a continuous flow, but actually they were particles. The duration of such falling down was also very short, only the order of ten seconds.
Fig. 14 shows the temperature distribution of the tube below the RPV wall. It should be first noted that the maximum temperature, which appeared at KP-21, 65 mm below the RPV inner sur-
M. Naitoh et al. / Annals of Nuclear Energy 118 (2018) 212–219
face, was less than 1000C and that the tube did not melt. Second, it should be noted that the temperatures showed large deviation even at the same elevation. The temperatures at KP-21 and KP23 were higher than ones at KP-22 and KP-24. The reverse trend was shown below KP-23 and KP-24. Based on this temperature behavior, it was estimated that the corium flow did not meet the flow path in the tube, but it flowed down as particles. Fig. 15 shows the temperature distribution in the RPV wall. Since the RPV wall received heat from the corium on the wall and at the same time released heat radially and axially, the temperature was higher at closer location to the inner surface of the RPV wall. The measured maximum temperature was 581C, which was well below the melting temperature of the steel. 5. Post test inspection Corium ingression into the tube was inspected. As stated, the corium was considered to flow down in the tube as particles. Some of them were caught on the T/C support clamps as shown in Fig. 16. The pictures were taken by a borescope which was inserted from the bottom open end of the pipe. They looked like a metal (steel). Nothing had deposited on the tube wall below the RPV wall. The tube inside the RPV had perfectly melted. The melt material solidified on the surface of the RPV and was hardly removed from the RPV wall. The test tube was welded at the inner surface of the RPV wall as shown in Fig. 2, and the test tube was anchored to the RPV at the weld portion. The weld portion was covered by the solidified material and was integrated with it, resulting in one large ingot. Some melt flowed into the tube and solidified, resulting in blockage of the flow path in the tube. The depth of the melt ingression from the inner surface of the RPV wall was 60 mm at the annulus region and 65 mm at the inner tube as shown in Fig. 17. Several samples were taken from the solidified materials; solidified material on the RPV inner surface, bottom of the ingot in the inner tube, and released materials from the annulus and dropped on the support for the T/C wires. The samples were analyzed by using ICP-AES (Inductively Coupled Plasma-Atomic Emission Spectrometry). The chemical analysis result showed very small amount of uranium, less than 0.016 wt% (0.16 mg/g), and ferrous component of more than 70 wt% for all the samples. It was concluded from the chemical analysis that the stainless steel tube melted in the lower plenum and some of the melted steel dropped down through the bottom of the tube and the material which irrupted into the tube soon solidified and blocked the tube preventing further corium ingression. 6. Conclusion A melt behavior of a BWR in-core monitor guide tube (ICM-GT) for a short range monitor (SRM) and an intermediate range monitor (IRM), which penetrates a RPV bottom wall, under a severe accident condition was tested using a corium with a mixture of UO2, ZrO2 and Zr. The test specimen was a full scale SRM/IRM guide tube with the same dimension as one of Fukushima Daiichi Unit-1 but short length. The melt corium was generated in the upper crucible and it was discharged into the lower crucible where the test specimen was installed. The test result showed that:
219
(1) When the corium was discharged into the lower crucible, the guide tube inside the RPV was first attacked by the corium and then it accumulated in the lower crucible. (2) Due to the first attack of corium, the guide tube in the lower crucible melted from the top. (3) Some of the melts were caught on the T/C support clamps as particulates. (4) Small amount of melts had fallen out from the bottom open end of the tube into the pedestal as particles, and such outflow stopped soon because a stainless steel melt had solidified in the tube and blocked the flow path. (5) The depth of the melt ingression into the guide tube was 60– 65 mm from the inner surface of the RPV bottom wall, where the tube was completely blocked by the solidification of steel melt and no oxide melts had ingressed there. In the test, the time of induction heating to the lower crucible was limited to only about 23 min. At the actual plant, on the other hand, the high temperature corium with decay heat generation might stay longer in the lower plenum resulting in heating the RPV bottom wall. In such a case at the actual plant, the depth of the melt (steel melt) ingression into the tube might become much longer, but the possibility of ingression of oxide melts was considered to be small because the tube might be blocked by the solidification of steel melt due to heat radiation from the tube outside the RPV. Moreover, the guide tube ejection from the RPV wall might occur due to melting of the weld portion at the inner surface of the RPV by heat conduction from the high temperature corium. 7. Future work This paper describes only the test result. The important result obtained from the test was that the tube was fully blocked by the ingression of the melt of stainless steel guide tube and its solidification. At the actual plant, detailed damage behavior of the guide tube can be evaluated by heat conduction calculation from the high temperature fuel corium with decay heat generation. The accident progression analysis of the actual plant shall be conducted considering such heat conduction calculation. Acknowledgements The test has been jointly conducted by the Institute of Applied Energy and Korea Atomic Energy Research Institute with a financial sponsorship of the Japanese Ministry of Economy, Trade and Industry. The test specimen was provided by Hitachi-GE Nuclear Energy, Ltd. The authors would like to thank Dr. Jaebong Lee of KAERI for the valuable discussions. The authors would also like to thank test engineers of KAERI for the implementation of the tests. References Hong, S.W. et al., 2003. Application of cold crucible for melting of UO2/ZrO2 mixture. Mater. Sci. Eng., 297–303 Takasaki, Fumihiko, Mizokami, Shinya, Nagano, Mamoru, 2015. Inspection of the Damaged Reactor at the Fukushima Daiichi with the Cosmic Muon Radiography Muon Investigation Report, 2015 July (In Japanese). IRID (International Research Institute for Nuclear Decommissioning).