Japan Earthquake: nuclear plants Fukushima part 2

  • #1,716
Science Advisor
Ordinary house fires with modern plastic and wood furniture etc items easily reach "few hundred degrees" and more at ceiling level.
Usually that kind of fire won't heat up the whole concrete panel to that temperature. In such cases only surface spalling (few cm depth) is expected.

I found https://www.researchgate.net/publication/261436338_Fire_Spalling_of_Concrete_Theoretical_and_Experimental_Studies/download some time ago. At the level we need here I think the examples and descriptions alone will do.
  • #1,717
Charles Smalls
It is speculated that the high temperature molten fuel that fell to the bottom of the reactor caused some of the surrounding concrete to "melt".[...] Now they worry that some of the molten fuels might not be located in the limited area right under the reaction vessel, but also spread to a certain degree in the surrounding areas too.
I talked about it back in 2017, the process is called spalling. Great visual demonstration of it here:

and in depth explanation here. Basically extreme heat causes moisture and air pocket pores trapped inside the concrete to expand into steam which explodes the surface concrete off. Varying constituents in the concrete mix also expand from different rates which cause further degrading of the material. The presence moisture accelerates the process by facilitating the migration of water deeper into the concrete where it then rapidly expands as steam. This cycle then repeats.

The spalling process combined with potential damage from the initial earthquake are the most likely culprits behind the compromise in containment that have been causing the persistent ground water contamination problems.

P.S. Would be very helpful if we had a map of the rover location during that video. At certain parts the amount of visual artifacts and grain caused by radiation is off the charts
  • Like
Likes Sotan, russ_watters and artis
  • #1,718
(not sure how to quote only part of the post above)

P.S. Would be very helpful if we had a map of the rover location during that video. At certain parts the amount of visual artifacts and grain caused by radiation is off the charts

Maybe this information will help a little:

- There are some still camera shots on this page . They might help figure out how things look in various areas. The descriptions and photos can be identified easily based on time stamps:

Pedestal opening, bottom, left side (21:19:25) / Pedestal opening, bottom, right side (21:19:37)
Pedestal opening, bottom, right side, area above the deposit (22:04:54) / Pedestal opening, left side, are above and below the deposit (22:04:57)
Pedestal opening, bottom, right side, area under the deposit / Pedestal opening, inside, in front, looking down (22:10:37)
Pedestal opening, inside, in front (1) (21:21:47) / Pedestal opening, inside, in front (1) (21:20:24)
Jet deflector D, looking down (11:43:50) / Jet deflector C, looking down (15:50:47)
Jet deflector D, back side (13:48:22)

- If you play this movie from the same page linked above, you have these explanation at the bottom at these times (there are 4 cameras, of which the main one, looking forward ("front") is the one providing the image at bottom right. The others are labeled "side/under" - this is the one at top left; "front (in air)", top right; and "rear" (bottom left).
00:21 Jet deflector F, looking down (or: seen from above)
00:28 Jet deflector E, looking down
00:40 Deposit on the floor of the containment vessel
00:56 Vicinity of the "machinery drain sump pump" (? installed in the dry well?)
01:06 Pedestal basis/bottom/foundation area
01:28 Front side of jet deflector E
01:41 Back side of jet deflector E
01:50 Reactor recirculation system (A) pipes
01:57 In front of the opening of the pedestal
02:07 Pedestal basis/bottom/foundation area
02:19 Jet deflector E area
02:34 Jet deflector D, looking down
02:45 Back side of jet deflector D
02:56 Jet deflector C, looking down
03:07 Pedestal basis/bottom/foundation area
03:17 Area inside pedestal, at front, (probably) seen through the opening
03:34 Pedestal opening, basis/bottom/foundation area
03:46 Pedestal opening, basis/bottom/foundation area and inside/at the front

- I may have not used the correct terms every time. I don't know what the jet deflectors are.
- Looking at the screen, sometimes the cameras seem to witness different amounts of radiation artefacts. Maybe they are different cameras or maybe the direction they are pointing at matters.

- On the same page linked at the top of this message there is also a link to this document. Below is a translation of the summary (google translate and a little checking). There is more information in the document, I don't have time right now to translate it all, but if anyone is interesting in something in there, let me know. It explains in more detail and using a map the location of various photos. The "Pedestal opening" which is examined in detail is depicted on the maps by a red line square and the red letters "ペデスタル開口部".

Fukushima Daiichi Power Station Unit 1 Containment vessel internal survey Completion of detailed visual survey of the outer periphery of the pedestal by ROV-A2
<Reference Material>
 A PCV internal survey was conducted with the aim of examining the recovery process and equipment for recovery of deposits in the containment vessel (hereinafter referred to as PCV).
 In the survey, a remote-controlled robot (hereinafter referred to as “water ROV”) developed for this purpose was used to carry out “detailed visual survey inside and outside the pedestal * 1” and “sediment thickness measurement” in sequence. We are planning to get information such as quantity and origin of the sediment.
 Prior to these surveys, ROV-A installed a "guide ring * 2" inside the PCV from August 8 to October.
 The “Detailed visual survey of the outer periphery of the pedestal” using the ROV-A2 was conducted from March 14th, and resumed from May 17th after a temporary suspension of the survey. On May 17th and 18th, we checked the condition of existing structures and the spread of sediments near the pedestal foundation and near the jet deflector * 3 (E) (slide 3 frame).
 On May 19th, near the pedestal opening and near the jet deflectors (C, D) (slide 3 green frame), the state of existing structures and sediments was investigated. We confirmed the spread situation. (Refer to slides 4 to 7 for details.)
 On May 20, 21st, neutral bundle measurement was performed near the pedestal opening and near the jet deflector (E, F, H) (slide 3 small frame). We are currently conducting analysis and evaluation of the obtained data, and will inform you of the results of neutral bundle measurement after the analysis and evaluation are completed.
 The lifting of ROV-A2 during the day started from 11:00 am on May 23, and was completed at 3:23 pm. This completes a detailed visual survey of the outer periphery of the pedestal. After the resumption of the survey on May 17, no significant change in the PCV position and water turbidity was confirmed, and there was no impact on the survey. In addition, no defects were found in the camera image mounted on the ROVA2 during the water.
 We will evaluate the situation confirmed in the detailed visual survey in the future, and if necessary, we will carry out the survey in the mid-season ROV-A2, B, C, D, E scheduled to be carried out in the future. In addition, the results obtained by neutral bundle measurement will be utilized in the future to narrow down the survey range of "sediment debris detection (nuclide analysis / neutral bundle measurement)" by ROV-D.  In the investigation, the work was carried out after constructing a boundary * 4 so that the gas in the PCV would not leak to the outside, and no significant change was confirmed in the data of the monitoring post and dust monitor, and the plant parameters. No radiation effects on the surrounding environment have occurred. We will continue to carefully investigate, safety as our top priority.
  • Like
Likes Charles Smalls
  • #1,719
Asahi Shimbun article of May 28
"Photos taken by a remote-controlled robotic device sent into the No. 1 reactor found that a large portion of the concrete base supporting the pressure vessel appears to have melted, leaving only a metal framework holding up the pressure vessel.

Experts are now saying the remaining structure may not be strong enough to withstand a big earthquake, a troubling prediction given that the region has been hit by a number of strong temblors in recent months."
  • #1,720
Charles Smalls
Hey all, I was going over some old posts today and was saddened to realize that Jim Hardy had passed away. I remember him being popular here and I'm sure he'll be much missed.

My initial reason for going back was to recall exactly where things had left off and what the main unanswered questions were. With everything back in focus, I must say I find the latest results slightly disturbing. Back in 2017 NHK aired a fantastic investigation into the 3 units which @Sotan posted. Sadly no longer viewable on youtube but we have most of the more useful images still saved here in post #1409. Among the things they theorised, one was possible rapid melt out of fuel from RPVs via the bottom head drainline. Resident BWR expert @Hiddencamper actually made quiet mention of this potential weak point prior in post #1373. This gave a plausible method for rapid core-exit focused into one area without CRD damage, and suggested the type of spalling we should see.

you can clearly make out the textbook morphology of heat induced damage where the material breaks up in a characteristic layered sheet-like manner. But it's just an example. The real spall site of interest is directly outside the Unit 1 pedestal doorway to the PVC. This is the presumed pathway any molten fuel would have flowed along as it left the RV and the latest data shows that the deposited material found there is too inactive to be the fuel itself. That indicates that the fuel did pass through or under that area.

5 years later we finally see visual confirmation of the very heavy spalling we suspected.

Also back in 2017, TEPCO released a very interesting sequence of videos filmed in Unit 5. (right-hand links still work). There's something eerie about the sequence as they basically seem to be charting the exact pathway the fuel would have taken after escaping the the RPV in the sister units. From the CRD room above, through the grating to the pedestal floor and then out the doorway into the PVC proper. This is exactly what the latest 23/05/2022 footage covers and it definitely helps to compare the "before" and "after". One question very rarely raised, is where escaped fuel proceeds from here. Especially given it's super-heated molten state, the spalling, and the channeling effect of the pedestal opening. So let's talk about Jet Deflectors (JD). The PVC drywell (DW) is connected to the tours wetwell (WW) by 8 large vent pipes. These pipes are capped by open vented covers on the DW end called jet deflectors, and permit the movement of air between the DW and WW to prevent damage from potential pressure differentials between the two vessels.


In TEPCOs pdf they are coloured yellow and labeled A-H.


And there's a photograph of one in normal condition in the background here labeled X-5A

Back to the 2017 NHK documentary:

(34:03) According to this graphic, they suspect hot molten fuel somehow made it's way into the suppression pool and burned through the bottom of the torus to contaminate and leak water from there.


(34:09) They seem to be experimenting with different concrete consistencies and mixtures so they can get a mix to pipe into the suppression and plug the suspected leaks.

There was speculation on the thread in 2017 on exactly what this image showed. Clearly a suspected problematic leak in the torus they hoped to plug, but the bright orange colour they used to depict it was unclear. It may have been representing fuel, but at the time we concluded it was simply using a 'hot' colour by coincidence to highlight a suspected leak. If we go to TEPCOs new 23/05/2022 ROV video, we see a lot which may actually be relevant to this. Ignoring the severe pedestal spall damage for a moment, look bottom right at 2:12 and 2:20. Distinct layers where a 'crust' has formed above a material no longer present, leaving a void. The material has since headed off toward the JDs and left this curious gap or hollow space behind. It looks for all the world like a https://caves.org/virtual/virtual_tube/birth1.jpg, where molten material has passed on, leaving a thin crust of material 'floating' above.


Impressively, the ROV follows these crusts and material build-ups which have pilled up against the JDs, even going so far as to enter JD-D to film the other side (2:46). The high amount of noise and sparkle indicates the presence of high radioactivity, but more tellingly, notice that in JD-C, D and E, the video is edited to cut just before the camera pans round to film vent pipe and pathway down to the WW (2:54). In an earlier ROV mission, we see the full pan around of JD-G, but none in these three. As JD-C,D and E are pretty much directly opposite the pedestal opening, I speculate that the molten fuel would have flowed and spalled its way straight into these vents, causing a large amount of severe damage down to the WW which TEPCO don't wish to publicly disclose at this point. It's also possible the experts who consulted with NHK to produce that 2017 special suspected this all along (wish it was still available). That would align well with those graphics. What all this means as far as the 7m thick basemat, with the fuel likely taking a shortcut around it to down under the WW, and whether this is also the case in the other 2 units is still unknown. Probably not in unit 3 as the it still holds water quite well, but very likely in unit 2. It's difficult to build a complete picture as TEPCO is still being selective with the images it discloses, but I think this confirms the most likely route for possible ground penetration leading to the persistent ground water contamination issues they've been battling. Difficult to see how best they recover fuel that's gone as far as to leave the containment entirely. Hopefully we get a similar ROV investigation of Unit 3 soon, and some numbers of this one to estimate likely presence of local fuel to see exactly how much went where.

Take care all,
Rip old Jim
Last edited:
  • #1,721
RIP Jim Hardy...

Thank you Charles Small for your post. I cannot say anything of substance about the possibility of molten fuel going through the big pipes into the S/C. Animations at the time (see the 2018 version of that video below) tend to depict the molten fuel at the bottom of the pedestal but somehow not rising high enough to reach the jet deflectors. We know now from rover images that it did rise, deposits are shown engulfing the jet deflectors at least in part (and there is a scene somewhere in which someone estimates the thickness of the deposit at ~60cm, so it depends how high those jet deflectors are situated, above the floor). With the violence of the phenomena going on in the pedestal area at the time it is not impossible to imagine chunks/splashes of stuff going in all directions available. (Let me also mention here this article of Nov 2021 - Analysis of Fukushima-Daiichi Nuclear Power Plant Unit 3 pressure data and obtained insights on accident progression behavior - which may be of interested to some.)

- So indeed the 2017 movie you mentioned is gone private but it seems NHK keeps refreshing that with new information every year. there should be a 2021 version but it seems to be pay per view, only on NHK site. I was able to find the 2018 version of that video, which has interesting images and information, too. It is in Japanese but closed captions help, to a degree.
At 6:50 they imagine the fuel melting process, the molten stuff goes at the bottom and stays put. Doesn't rise too far up...
8:47: the handle of a fuel bundle seen in the solidified deposit
9:20 estimation of the thickness/height of the deposit, about knee-high
16:16 and after - images from Reactor 5, which is similar to Reactor 2 which was depicted until now. Workers go under the RPV, we see the control rods.
A lot of discussion, experts and workers' opinions, including comparison with Three Miles Island (where things were nowhere as difficult), and realization of the difficulties of the task of decommissioning.

- A similar video but more recent (Feb. 21 but includes images from 2018 too) is here. Again at 5:56 they go under the PCV of Unit 5, then go lower into the pedestal and at 7:50 talk about the thickness/height of the solidified deposit in Unit 2, estimating it at about 60 cm. *My thoughts seeing this: Is it high enough to support the idea of molten fuel going through the big pipes into the torus? Is it a case of slow and quiet rising up of a liquid mass, like water in a sink when the faucet is left open and the bottom hole blocked, and then starts pouring out through these pipes - or is it rather a violent phenomenon in which everything flies everywhere and in that case we don't have to think too much about heights and levels...


  • p1.png
    151.3 KB · Views: 37
  • p2.png
    86.2 KB · Views: 33
  • #1,722
Staff Emeritus
Science Advisor
"Photos taken by a remote-controlled robotic device sent into the No. 1 reactor found that a large portion of the concrete base supporting the pressure vessel appears to have melted, leaving only a metal framework holding up the pressure vessel.
That's a puzzling statement given that the steel would melt at a lower temperature than concrete, i.e., if the concrete had melted, the steel should have melted before the concrete. Perhaps the concrete reacted chemically with seawater and whatever compounds were in the seawater.
  • #1,723
That expression - and those extraordinary images - puzzled me too, hence why I posted "does concrete melt in such a way that steel bars remain?". And was wondering of some chemical reaction too.

But now I think it's rather a matter of "lost in translation". In an interview one worker says "it melted... or evaporated" (!). So I think what they intend to convey is more like "that concrete... disappeared". It's not there anymore. They don't really explain how it happened, just describe the result.

This still leaves the question what happened exactly. But salt water, and I'm only talking intuitively, appears too benign to me to be able to cause such huge results (maybe I'm wrong). I am inclined to give credit to the spalling explanation that has been suggested. It's just that the spalling video and articles seem to speak of a more superficial or partial erosions of concrete, and again intuition tends to say what we see here is different. But maybe higher temperature, long contact time and the contact with a molten mass of metal (rather than hot air from a fire) over a long time kept the spalling process going...
Still - how the concrete got cleaned so nicely away from the steel, I mean it's eery. So temperature did not exceed melting point of steel. How thick was that wall. Is the concrete gone everywhere, 360 degrees? Or just in portions? Is the whole thing now really supported by those steel bars only? High temperatures would not have weakened that steel dangerously?

I suppose that is why they plan to continue investigations for the next 6 months.
  • #1,724
March 23 Report on measurements of the thickness of the deposit at the bottom of Unit 1 PCV
(The report is in Japanese. Page 3 contains most of the information.)
They used a version of the robot that floats on the surface of the PCV water and sends ultrasounds towards the deposit and analyzes the reflection. They thus estimated at 1.0 ~ 1.2m the distance from the water surface to the deposit surface; knowing that water surface is at 2m from the bottom of the PCV, the thickness of the deposit has been evaluated at 0.8 ~ 1.0m.
This robot cannot give detailed information about the structure of the deposit, but results indicate that at the top there is a layer of a couple/few centimeters of finer particles (like a powder or mud layer) and then a more solid deposit (pebble/rock type). Another robot will follow investigating that in more detail, and then another will try to get a sample.
  • #1,725
Science Advisor
Is the whole thing now really supported by those steel bars only? High temperatures would not have weakened that steel dangerously?
Rebars does not supposed to have (significant) structural strength without the support of concrete. High temperatures are just extra.
However: right now the mass of the supported structure has decreased by a lot (all fuel, water and most of the internal structure left the PCV...) so I hope/think it'll be temporarily OK with hanging on the connected pipes and with reduced support.
  • #1,726
Charles Smalls
How the concrete got cleaned so nicely away from the steel? Temperature did not exceed melting point of steel. How thick was that wall? Is the concrete gone everywhere, 360 degrees? Or just in portions? Is the whole thing now really supported by those steel bars only?
According to areas marked red in this TEPCO pdf, they seem to be estimating severe concrete damage to ~27% of the pedestal concentrated around the doorway. Plus the pedestal inner walls.

The pdf source says 2016 though, so it's possible that further damage has been observed in later investigations. Prior to any damage, the pedestal wall was approximately ~0.65m thick. That number comes from a 1989 paper on "Liner Failure Potential in MARK-1", which looks specifically at fuel attacking concrete and melting through the outer containment. The reactor they base their study on is almost identical to Fukushima unit 1.

Couple of interesting things:

These values do not account for the sensible heat needed to raise the concrete components from the decomposition temperature to the melt temperature [...] Of particular significance is the high melting temperature of the limestone concrete, which is higher than the melting temperature of steel

So there's a range of temperatures at play here: one where concrete undergoes decomposition or ablation (what I called spalling), a higher one where steel will melt and a higher one yet where the concrete itself becomes liquid. (depending on concrete type). This is one possible explanation for why unit 1 can have such dramatically damaged concrete while still leaving the steel rebar unmelted.

The gases rising from the decomposed concrete( H20 and C02 ) cam react with the zirconium metal present, leading to a heat source that can overshadow the decay heat [of the molten fuel].

So depending on the mix and composition of the meltout, the zircalloy cladding within it can actually accelerate the decomposition once it contacts the concrete. So that's a nice little positive feedback loop there.

> [Our] spreading analysis ignores the possible presence of water in and out of the pedestal. Such a dry case is expected to pose a much higher threat to the liner than the wet case, due to the considerable potential for melt quench and cooldown via thermal interaction with the water pool.

This lines up well with the images from @Sotan's news investigations.


I haven't seen the underlying data, but all the event reconstructions I'm aware of all specifically depict unit 3 as having water in the containment prior to fuel melt out. This makes sense when we look at the three units today with unit 3 holding water best, while units 1 and 2 struggle, suggesting possible containment penetration via fuel melt. Difficult to say with TEPCO's highly selective data release. I know I mentioned this before but I think it warrants restating:

In the first investigation video at 2:41 the ROV enters jet deflector H furthest from the pedestal doorway, and we see a nice clear view as the camera pans around and shows a relatively clean vent tube.

Later in part two, at 1:43 the ROV enters jet deflector E directly opposite the doorway. The timestamp is 14:17 and the camera just begins to glimpse what look like fuel deposits. Scandalously the video cuts, picking up again fully 3 hours at 17:05. At 2:44 of the same video, the ROV next enters jet deflector D. Timestamp is 13:43 and there is a high amount of radiation noise and sparkle. The camera pans around barely 1/4 of the way showing some material intrusion, before abruptly cutting off and picking back up again outside the deflector with the timestamp 15:50. Incredible.

I'm sure we'll get more interesting releases from TEPCO in the future but whether they can be taken at face value or require extra work to interpret is another matter. Some things however, like the danger of that pedestal and it's earthquake fragility can't be obscured.
  • #1,727
Some information from this document dated 27 Oct 2022. (Page numbers indicated below are as given in the top left corner by Adobe Reader for whole PDF file (total 57 pages). There are other page numbers too, for each section, and this can cause some confusion.)

P4: They are planning to sample the water contained in the S/C chamber of Unit1 PCV. P5: One of the means to improve quake resistance of Unit 1 PCV is lowering the level of water in the PCV. They are therefore working on installing the necessary pipes, while making use of existing pipes too (like CUW – reactor water cleanup system?). Note 1* The needed pipes for removing such water in Unit 3 – making use of existing RHR, residual heat remover, pipes too – are already installed. As a first step they need to check the properties of the water in the S/C, and for this they plan an inspection/sampling operation, which has started these days and should end in Jan next year. There are some diagrams and a photo explaining the concerned are: first floor of Unit 1 reactor building with candidate spot for inserting new pipes (1-10mSv/h area); existing pipes; non-return valve on the CUW pipe. P11: Process planned for lowering the level of water in the PCV of Unit 1. The approximate goal, for now, is to bring water level around the “center” of the S/C pipes.

P12~33 show the work being done in preparation for new inspections and, later on, sampling of debris from Unit 2 RPV. There is a lot of site preparation, robot testing and improving, and personnel training going on.

P34~ : New data presented by IRID / JAEA regarding the characterization of the fuel debris situation inside the damaged RPVs. As they keep gathering information, they work on improving the degree of understanding regarding what exactly is in there.

P36: refers to Unit 2. The new findings are at the bottom of the page. – As the fuel debris accumulated at the bottom of the RPV, a new rise in temperature greatly affected the course of the accident; the fuel debris containing a lot of metal melted first, at 1000-1300°C, then damaged the RPV wall in places, and then poured onto the pedestal floor. – The fuel debris that leaked/accumulated on the pedestal floor got cooler and did not succeed in “melting” the concrete; therefore, if right after the accident it was assumed that a MCCI (Molten core/Concrete Interaction?) happened, it is (now) assumed that the MCCI was almost absent.

P37: refers to Unit 3. The new findings are at the bottom of the page. – As the fuel debris accumulated at the bottom of the RPV, a new rise in temperature greatly affected the course of the accident; the fuel debris containing a lot of metal melted first, at 1000-1300°C, but then, the fuel debris containing a lot of oxide compounds reached a high viscosity state (like a “hard liquid mixture”? sorry can’t find a good word), at 2000-2300°C, which damaged the RPV bottom on a rather large area and, over a few hours, fell on the pedestal floor. – The debris thus accumulated on the pedestal floor, due to its high viscosity, did not closely interact with the concrete; therefore, if right after the accident it was assumed that a MCCI (Molten core/Concrete Interaction?) happened, it is (now) assumed that the MCCI was almost absent.

P38: new insights on the state of the fuel debris. They speak about results of analysis of some small grains/particle harvested in some previous investigation, results which are now added into a “debrisWiki” database. I think it is accessible here https://fdada-plus.info/wiki/index.php, English version here: https://fdada--plus-info.translate...._sl=ja&_x_tr_tl=en&_x_tr_hl=ja&_x_tr_pto=wapp

P43~ : Report on the second part of the investigation into Unit 1 PCV. Interesting new images on P48-50, including a weird ball-shaped piece of debris found on the floor of the pedestal. Looks like a solidified bubble with holes in it.
  • Like
Likes SteveElbows, Gary7 and Rive
  • #1,728
Thanks for that update. The DebrisWiki site is useful. They also review their understanding of core damage and accident progression on that site. As we already know from past stuff such as the missing concrete in unit 1, there is an assumption of plenty of MCCI there. And just like in the report you mention, they now think that MCCI didnt happen much in reactor 2. When it comes to reactor 3, I think perhaps they are still hedging their bets a bit. eg this is the translation from DebrisWiki In these respects:

Unit 1:

It is highly likely that almost all of the fuel debris migrated to the pedestal, spread over the pedestal floor, caused Link Removed , and eroded the concrete.
The causes of hydrogen generation include the cooling water-Zr reaction and MCCI in the core, but it is highly likely that most of the hydrogen generated by the former has already been discharged from the reactor building. The following two routes are conceivable as hydrogen release routes.

  • Hydrogen generated in the core → safety relief valve (SRV) → S/C → D/W → PCV top flange → reactor building
  • Hydrogen generated in MCCI→D/W→PCV top flange→Reactor building
After the high-temperature fuel debris migrated to the pedestal, it is possible that sufficient cooling water was not injected into the pedestal for about 10 days. Therefore, the pedestal is filled with high-temperature steam (approximately 400°C according to the rough analysis using the STAR-CCM+ code [1] ), and the inner wall of the pedestal is heated by radiation (the rough analysis using the STAR-CCM+ code [1] ] estimated that the temperature was approximately 800°C), and some of the concrete and other structural materials may have been damaged at this stage.

Unit 2:

It is highly likely that the transition from the lower plenum to the pedestal consisted mainly of molten metal or a relatively low melting point such as U-Zr-O. Such migration of core material to the pedestal occurred within a period of several hours, and it is possible that most of the oxide-based debris remained inside the RPV. At the lower end of the CRD near the application in the RPV, deposits thought to be debris were found, and it is possible that a small amount of debris was released through this part. The gratings of the platform directly below this position and the intermediate frame below it have fallen off, but the grating of the adjacent part of the intermediate frame (the grating of the platform directly above remains) has also fallen off. In addition, considering the condition of the deposits remaining on the grating that remained on the platform, it is presumed that the spilled debris was a highly viscous fluid with a temperature below the melting point of steel. Since the temperature of outflow debris was not so high and the amount of outflow was small, it is considered that MCCI hardly occurred.

Unit 3:

Since the D/W pressure began to rise from around 0:00 on March 14, it is believed that the core material that had migrated to the pedestal at this point was evaporating the cooling water that had been in the pedestal. However, since the D/W pressure rise was slow, it is highly possible that the migration of core materials to the pedestal was extremely slow. After 0:00 on March 14, the D/W pressure gradually increased until around 7:00 on the same day. This pressure increase is thought to correspond to the steam generation at the pedestal, and it took about 7 hours. It is presumed that pedestal migration of debris has occurred on the scale.

Investigations inside the pedestal confirmed deposits of 2 to 3 m, but this volume cannot be explained by the fact that the core materials are densely packed, and there are considerable gaps inside the deposits. There is a possibility that If a significant MCCI had occurred, the debris would have reached a considerably high temperature, which is incompatible with having many gaps in the sediments. Based on this, it has been pointed out that MCCI may not occur very often in Unit 3. On the other hand, it is possible that remelting occurred inside the pedestal sediment, and in that case, the phase state is expected to become complicated due to component segregation, remelting, and resolidification. Again, the solidification rate is likely to affect the properties of fuel debris.

I got those quotes from the three pages that are linked to from this page about accident progression:


The visual diagrams on those pages, which are mostly the same as the ones in the report you mention, also imply they might be hedging their bets in regards unit 3 MCCI, since there is still some colouring down in the pedestal floor concrete area of that reactor.

Those summaries are quite useful in other ways. It had been years since I last looked at this sort of detail. So it was useful to see them putting the timeline of observed reactor and D/W pressures into a short narrative about the meltdowns that was not trying to obfuscate the seriousness of the incidents and extent of fuel melt like some of the early 2011 analysis did. They even mention the neutron detection episodes in an attempt to make educated guesses about timing of various stages of fuel damage and debris slumping at reactors 2 and 3.
  • #1,729
By the way, after quite a number of years of not paying as much attention as I did in the first years after the disaster, I found it rather tedious trying to search for the English versions of TEPCO documents that come out these days, due to changes in how their website is organised. Sometimes I was still finding the documents via links they put on the photos page that still work, but these tended to be initial reports of that days work and the photos and video the work yielded, rather than final analysis. I know that many of these investigations tend to be reported here in this thread thanks to the hard work of people here offering translation of the Japanese versions, but sometimes I want to look at English versions later on to help keep everything organised in my mind.

Anyway after much stumbling around today I finally found the page where the interesting English reports are announced and linked to these days (as well as plenty of less interesting ones which I skip over). https://www.tepco.co.jp/en/hd/decommission/information/newsrelease/reference/index-e.html

A couple of my favourites from the last 4 years or so:

Unit 1 internal investigation report from late May 2022:


The April 2018 Unit 2 investigation document which remains one of the better ones I've ever seen from them, in terms of pedestal area analysis with clear photos and useful explanations of the location and the detail:

  • #1,730
Thank you very much SteveElbows for the two very informative and well written posts. It's been a while so let me say again that due to my lack of time (but more importantly lack of... proper knowledge in the field) my posts are at best approximative. I'm just hoping they can restart the discussion :)

The page you found with English reports is great, it is a little delayed of course but at this point such a delay is probably not a big problem.

More on topic: they seem eager and pleased to state that evidence suggests not much MCCI happened - but how does that match the images with the naked steel bars supporting the RPV?
  • Like
Likes SteveElbows
  • #1,731
Thanks. I am just a layperson too, who had a very large amount of time to pay attention to this disaster in the first years, and who always enjoyed talking here about it 10+ years ago with knowledgable and inquisitive people. And now its only in recent years that they reached the point of site investigation that renewed my interest, eg some reasonably good quality images of the state of things in the pedestal areas.

In regards their comments about MCCI and how that relates to eroded concrete images, I don't think there is a contradiction. Because the photos showing the exposed rebar are from unit 1, and they do think MCCI happened at that particular reactors base. Its unit 2 where they downplay the possibility of much MCCI, and then with unit 3 they are hedging their bets about the extent to which it might have happened.
  • #1,732
The Nuclear Regulatory Authority (NRA) of Japan also has periodic meetings in which experts from various institutions and organizations discuss recent reports coming from Fukushima Daiichi, as well as additional studies and results. Unfortunately they are mostly in Japanese, but exploring the English version can also provide some data.

- The list of NRA meetings dedicated on the analysis of the Fukushima Daiichi accident is on this page. Most recent was on Oct 31. The blue rectangle named 会議資料 leads to the written materials used in that particular meeting. Lower on the page there is a link named 会議映像 which leads to the recording of the meeting on youtube. The recording for the meeting of Oct 31 is here, it is over 4h long, and one of the subjects discussed (among many) is the matter of the missing concrete and exposed rebar. At around 38:19 an associate professor from Osaka University (mr. Oishi) presents a study carried out by his team, discussing this subject. They make some theoretical analysis of the factors that might have led to that state of the bottom of the RPV and in the second part they present some experimental results on concrete and rebar behavior at high temperatures.

I am picking to report this not because there is something extraordinarily new or significant to report but because in their analysis they mention a second factor, besides MCCI, that might have contributed significantly to the disappearance of the concrete. At 42:04 they list the possible factors and two of them (marked with A in the rightmost column) are considered significant, one (B) is so-so, and 4 others (marked with C are considered less likely to have contributed. A-factors are MCCI and "dissolution in water": some components of concrete can get dissolved in water, especially SiO2. The reaction obviously needs the presence of water and can happen even at temperature as low as 200oC. (The B-factor is mechanical erosion.)

The chief of NRA at 1:22:19 considers this as significant. The hot molten fuel falling in the water at the bottom of the PCV inevitably raises the water temperature and the creates conditions for the dissolving of the concrete, while leaving the steel bars relatively unaffected - even without actual MCCI taking place. Or at least this possibility too should be taken into consideration and further investigated through actual experiments in the lab.

Of course this is just a very little snippet taken from a 4h meeting. So much to see and read...
  • #1,733
Staff Emeritus
Science Advisor
P37: refers to Unit 3. The new findings are at the bottom of the page. – As the fuel debris accumulated at the bottom of the RPV, a new rise in temperature greatly affected the course of the accident; the fuel debris containing a lot of metal melted first, at 1000-1300°C, but then, the fuel debris containing a lot of oxide compounds reached a high viscosity state (like a “hard liquid mixture”? sorry can’t find a good word), at 2000-2300°C, which damaged the RPV bottom on a rather large area and, over a few hours, fell on the pedestal floor.
I think "hard liquid mixture" simple means a very viscous material, perhaps shy of melting, where the material is not quite liquid, but has little strength/stiffness. In other words, in such a state, there is essentially no yield strength and the material flows, rather than creep. That usually a state of matter the mechanical/structural engineers would avoid.

The statement about the "fuel debris containing a lot of metal melted first, at 1000-1300°C," is interesting because I'm not sure what 'metals' or rather alloys, that would be, unless it refers to some lower melting eutectic/eutectoid or mix of such. Type 300 stainless steels (e.g., 304/316) would have melting points about 1370-1400°C (~2500-2550°F), nickel alloys (e.g., X750, Inconel 718) slightly higher, and Zircaloy (1820-1850°C, ~3310-3360°F), and UO2 melts at ~2840°C (~5070°F); however, corium, a combination of alloy and metal oxides could have a lower melting point depending on types of metals and stoichiometry. Certainly, beyond 1400°C, such a hot mass from the core could melt the steel supporting structure and flow downward throught the pressure vessel (bottom head) and onto the structures underneath, e.g., pedestal. This would suggest that they did not get cooling water into the system.
  • #1,734
Hello Astronuc. Let me add a few things to what you said above - mainly just to keep the topic going and because I had some time to browse more.

- About the lack of cooling water in the system, indeed there seem to be plenty of such mentions in the "Accident progression at Unit 3" which you can reach by visiting the DebrisWiki site linked by SteveElbows in his post of Nov 10 above - or via this link which gives it in google translation. "Water injection by HPCI stopped...boiled lower plenum water rapidly supplied steam to the core... depletion of liquid water in the lower plenum... which caused a rise in the temperature of the debris in the lower plenum... the RPV boundary was damaged and the debris flowed toward the pedestal... evaporating the cooling water that had been in the pedestal." (I highlighted just the phrases referring to the cooling water.)

- The above "Accident progression at Unit 3" text does not seem to clarify much my poor translation about "metals" melting. So I went back to Page 37 of that PDF to try and improve my translation. There was just one more mention that I overlooked: "Oxidized debris: from all the fuel debris (this refers mainly to) uranium, zirconium and stainless steel components in an oxidized state". As for my translation I cannot improve it much after all, but if I were to, it would be: "- As the fuel debris accumulated at the bottom of the RPV, a new rise in temperature greatly affected the course of the accident; the (parts of) fuel debris containing a lot of metal components melted first, at 1000-1300°C, but then, the (parts of) fuel debris containing a higher proportion of oxide compounds reached a high viscosity state (the literal translation would be “solid-liquid mixture”), at 2000-2300°C, which damaged the RPV bottom on a rather large area and, over a few hours, fell on the pedestal floor." I think what puzzles you a bit is the lower range of temperatures mentioned (1000-1300°C).

I tried to find more information about what type of compounds they suspect melted in that range. Among the documents shown in NRA meeting mentioned in my previous post there is this PDF which deals with findings of the PCV inspection of Unit 1. Second slide of this PDF (with a big number 37 at the bottom) mentions the amounts of various materials present in the PCV. Amount of fuel debris 279 tons of which 76 tons of fuel component (UO2...), 73 tons of fuel structural materials (Zr, SUS - I guess SUS refers to stainless steel), 130 tons of concrete. The volume of fuel debris is ~16m3. Then melting points are mentioned: ~2800°C for UO2, ~1850°C for Zr, 1400-1450°C for stainless steel and - only this one seems to be missing from your post - Pb, lead from shielding components, with a melting point of ~330°C. They don't say how much lead might have been in there so maybe not "that" much. Not sure if it matters much with regard to the more damaging processes that took place at higher temperatures.
- Next slide (number 3 or 38, whichever you like) mentions "melting of metallic components from inside the RPV - melting point somewhere around 1200-1300°C."

- Moving on to another subject, in the list of documents from that NRA meeting the documents which contain in their name 4-1, 4-2 and 4-3 talk about the observed significant deformation of the shield plugs. 4-1 reports that older finding that the shield plugs of Unit 1, 2, 3, 5 show a certain degree of deformation. Most recently Unit 6 was investigated and the results are shown in diagrams. The shield of Unit 6, some of its central parts especially, are deformed, as if they got lowered in some areas by about 20mm (page numbered 138-139). Page 140 shows diagrams of the shield plugs of Unit 2 and 5, with a similar phenomenon. Page 141 shows the shield plug in Shimane NPP and Tsuruga Unit 1, which are considered to show no signifivant deformation. Last page in the slide summarizes: Fukushima units 1, 2, 3, 5 show deformation of the shield plus; Fukushima Unit 6, Shimane 1 and Tsuruga 1 shield plugs show no deformation. Perhaps for Fukushima 1 and 3 the deformation can be linked to the hydrogen explosions but for Fukushima 2 and 5 it's unlikely that the deformation was caused by some external force. Possible causes may be: deformation already existing or occurred at construction/installation time; or gradually occurred in time. It may have played a role in creating a route or contributing to the escape of radioactive materials during the accident. This is further explored in documents 4-2 and 4-3.

- Finally - unrelated: a video from the May 2022 inspection of Unit 1 PCV which I only saw now for the first time, maybe someone is interested to see it. The exposed rebars, again... (By the way somewhere in the 4h video of the NRA meeting and in one of the documents they mention that there appears to be a sufficiently large route for then to some day guide one of those underwater robots into the pedestal.)
  • #1,735
Staff Emeritus
Science Advisor
https://www.marlinwire.com/blog/what-is-the-melting-point-of-stainless-steelPb, lead from shielding components, with a melting point of ~330°C.
I'm not sure where Pb would be used, even as shielding, inside the RPV.

In stainless steel, typically 300-series, e.g., 304/316, would melting temperatures from about 1375-1400°C for 316, and about 1400-1450°C for 304, depending on carbon and alloying content.
  • Grade 304. 1400-1450°C (2552-2642°F)
  • Grade 316. 1375-1400°C (2507-2552°F)
  • Grade 430. 1425-1510°C (2597-2750°F)
  • Grade 434. 1426-1510°C (2600-2750°F)
  • Grade 420. 1450-1510°C (2642-2750°F)
  • Grade 410. 1480-1530°C (2696-2786°F)
Ref: https://www.marlinwire.com/blog/what-is-the-melting-point-of-stainless-steel
Numbers can be found at BSSA.

Fuel components may be made of a 300 series stainless steel, and Inconel (or Ni-based) alloy (X750, 718 or 625), or Zr alloy (Zircaloy-4, ZIRLO, . . . ). In BWRs, the fuel cladding and water rods are composed of Zircaloy-2, similar to Zircaloy-4, but with a small amount (~0.05 to 0.07 wt%) Ni. The water rods and channels, the latter forming an envelope or shroud around the fuel assembly (designed to direct the water vertically and preclude cross-flow), may be formed of Zircaloy-4, Zircaloy-2, or NSF, which is similar to ZIRLO.

Ref: Zircaloy, https://www.atimaterials.com/Produc...zirconium/alloy/Zr_nuke_waste_disposal_v2.pdf

One does have to be careful of impurities, e.g., S and P, which may form low melting compounds in stainless steels, so those are kept very low, especially for stainless steels to be used withing the core. Stainless steels are typically used in cast, forged or otherwise worked form.
Last edited:
  • #1,736
- The WikiDebris site is growing nicely, I think we will use it more and more in the future.
Here is the article regarding "Melting and solidification behavior of fuels and structures during severe accidents" - in Google translation. I think it is more like a general review of the knowledge in the field, not necessarily limited to Fukushima, but very informative.

- Also it's been a while since IRID was mentioned. I found on their site an impressive 60-page PDF in English, recently posted, summarizing their numerous directions of research and development and the progress made in 10 years. You can find it here.
  • #1,737
Another nice batch of documents on the Tepco site that could be interesting for the specialists: a list of reports on unconfirmed and unexplained matters in the Fukushima Daiichi Nuclear Power Station accident.
The reports are in Japanese and simply pasting the URL in google translate does not work, but I discovered now that if you save the PDF and then bring the PDF file into google translate's field for translating "Documents", it will provide the English translation. (as always, small translation issues may be present; also might be limited to documents up to 300-pages).

The list contains the following items:
Overview of MAAP Code (617KB)
Topic List (3.05MB)
Latest analysis results by MAAP5 (1.54MB)
Investigation status related to estimation of the state inside the core and containment vessel (25.6MB)
Grasp of accident situation by sample analysis (3.75MB)
About the arrival time of the tsunami that hit the Fukushima Daiichi Nuclear Power Station (9.50MB)
Additional study on loss of emergency AC power due to tsunami (1.82MB)
Regarding setting of water injection amount in Unit 1 MAAP analysis (96.6KB)
Estimation based on the behavior of Unit 1 fuel range water level gauge (259KB)
Impact of the earthquake on Unit 1 (1.02MB)
Study on Reactor Water Injection by Fire Engine (1.01MB)
Study on water injection volume for Unit 1 reactor water injection by fire engines (528KB)
Estimation of accident progress based on Unit 1 measurement data and past analysis results (766KB)
Status of heat removal in Unit 1 isolation condenser (519KB)
Migration Behavior of Molten Fuel to Lower Core (3.63MB)
Estimation of the cause of high-dose contamination of Unit 1 RCW piping (1.19MB)
Analysis of the hydrogen explosion that occurred in the Unit 1 reactor building (2.37MB)
Estimation of accident progress at Unit 1 based on air dose rate monitoring data (1.33MB)
Identification of the cause of the high dose rate observed in the southeast area on the 1st floor of the Unit 1 Reactor Building (2.13MB)
Reactor pressure change in Unit 2 (310KB)
Containment vessel pressure change in Unit 2 (350KB)
Regarding setting of water injection amount in Unit 2 MAAP analysis (146KB)
RCIC flow rate after Unit 2 control power loss (313KB)
Status of the RHR system after the Unit 2 tsunami arrived (751KB)
Containment vessel pressure behavior from around 12:00 on the 14th of Unit 2 (928KB)
Relationship between neutrons observed at the time of the accident and fuel melting (294KB)
Investigation related to the integrity of the suppression chamber (S/C) of Unit 2 (689KB)
Evaluation of reactor pressure rise after forced reactor depressurization in Unit 2 using thermal-hydraulic analysis code (1.23MB)
Rapid increase in CAMS measurement values for Unit 2 on the 15th (153KB)
FP migration behavior estimated from the CAMS measurements taken on the 14th and 15th of Unit 2 (935KB)
Operation of SRV after core damage (932KB)
Changes in water level and temperature in the suppression chamber (S/C) of Unit 2 (1.36MB)
Estimation of reactor water level at the time when core damage/core melt progressed in Unit 2 (0.99MB)
Estimation of the reason why high dose rate was not observed in Unit 2 reactor auxiliary cooling water system (2.14MB)
Decrease in Unit 2 containment vessel pressure in the morning of March 15 (1.66MB)
Regarding the behavior of the S/C pressure gauge after 21:00 on March 14, Unit 2 (1.67MB)
Evaluation method for core damage ratio of Mark-I containment vessel (730KB)
Reactor pressure during operation of the high-pressure water injection system at Unit 3 (206KB)
Regarding setting of water injection amount in Unit 3 MAAP analysis (248KB)
Behavior of reactor pressure drop that occurred around 9:00 on March 13, Unit 3 (1.01MB)
Reactor pressure behavior from 02:00 to 12:00 on March 13, Unit 3 (1.08MB)
About the cause of the outage of Unit 3 RCIC (883KB)
Dose increase around March 20 (1.34MB)
Factors behind the increase in containment vessel pressure at Unit 3 from 3/11 to 3/12 (1.37MB)
Leakage from Unit 3 containment vessel and large amount of steam release (1.03MB)
Estimation of reactor water level behavior during the period when core damage/core melt progressed in Unit 3 (533KB)
Evaluation of the proportion of Unit 3 vent gas that flowed into the Unit 4 reactor building (1.09MB)
Study on Unit 3 suppression chamber water level (1.47MB)
Accident progress after depressurization of Unit 3 reactor (2.31MB)
Investigation of plant status during Unit 3 RCIC operation (1.72MB)
Last edited:

Suggested for: Japan Earthquake: nuclear plants Fukushima part 2

  • Last Post
  • Last Post
  • Last Post
  • Last Post