Japan Earthquake: Nuclear Plants at Fukushima Daiichi

[Luca Bevil]

Actually the problems with the contaminated water processing equipment are much more serious than anticipated.

http://www3.nhk.or.jp/daily/english/12_09.html

nothing to be happy or confident about

 

[westfield]

Early on some of the thermal images had scales attached to the photos and full scale was only about 5 degrees. If we can find the source of the photos there may be more information.

http://www.mod.go.jp/j/approach/defense/saigai/tohokuoki/temp.html"

http://i1185.photobucket.com/albums/z360/fukuwest/misc/Unit4Thermalinctable.jpg"

 

[tsutsuji]

Actually the problems with the contaminated water processing equipment are much more serious than anticipated.

http://www3.nhk.or.jp/daily/english/12_09.html

nothing to be happy or confident about

Only 0.5 tons per hour instead of the expected 12 tons per hour is flowing in one of the four units. A pipe is clogged or a valve is failing to open. The test run is postponed to tomorrow. The launch of the facility expected on June 17th or 18th : http://www.asahi.com/national/update/0612/TKY201106120141.html

 

[joewein]

The whole story is puzzling! I don't see any reason why they should have fuel in the RPV. But we see these hot spots. If it is water circulating between the RPV and the SFP why do we these delimited hot spots within the circle. Shouldn't the whole circle be an equally warm area?

No it shouldn't. You're looking at the warm water below through the remaining steel lattice of the roof structure of unit 4. The smaller and larger steel beams are cooler than the water below:

[PLAIN]http://img535.imageshack.us/img535/3165/gggss.png

Likewise, the purple zone between the red zone highlighted by the arrow and the circle is the crane structure above the fuel pool, which obstructs most of its thermal footprint. You tend to get sharper heat signatures at the edge of the building (arrow and bottom left corner), where steam escapes sideways through the blown out walls.

If you feel like studying more thermal images of unit 4, there's a whole bunch of them on the Ministry of Defence website, from March 20 to April 26:
http://www.mod.go.jp/j/approach/defense/saigai/tohokuoki/temp.html

Look for "第4号炉" for the unit 4 thermal images. On these images the pool area almost always comes out as the warmest part of the building, warmer than the reactor well in the center, just as expected.

 

[elektrownik]

On these images the pool area almost always comes out as the warmest part of the building, warmer than the reactor well in the center, just as expected.

NO, for example 4/20 sfp 18, core 29; 3/23 sfp 22, core 28;
We can see from scale that it temperature was bigger many times...

 

[ManuBZH]

Actually the problems with the contaminated water processing equipment are much more serious than anticipated.

http://www3.nhk.or.jp/daily/english/12_09.html

nothing to be happy or confident about

Nothing to be overly pessimistic about neither. Looking at the sheer size and complexity of the facility (and the incredibly fast construction time which is already a feat in itself), this is no surprise that they run into some problems.

We can only wait and hope that the whole thing will work as advertised and, once started, won't run into serious issues.

 

[joewein]

It sounds like they have problems with one of 4 units, each of which is designed to handle 300 t per day. If the other three units work as expected they would still be able to process 900 t per day, more than the 500 t they use for cooling each day.

It would be far more worrying if problems like these happened once they're using the system for processing highly contaminated water and not during trial runs with plain water or lowly contaminated water.

It's far preferable to encounter and solve the problems now rather than later.

 

[swl]

Think of a piece of iron in a fire. When you pull it out, you can look at it and from the glow you can tell instantly that one end is hotter than the other. However, unless you have some special experience, you probably don't have any clue at all exactly how hot the glowing end is and you don't really know if the non-glowing end is cool or not. All you know is that the glowing part is hotter than the not glowing part.

Unfortunately, the indicated temperature will be effected by the emissivity coefficient of each surface being measured. For example, if our iron bar in your example is polished on one end and rough finished on the other, there will be a huge difference in the IR emission detected. The emissivity coefficient will vary with both the surface finish and the type of material.

 

[etudiant]

It sounds like they have problems with one of 4 units, each of which is designed to handle 300 t per day. If the other three units work as expected they would still be able to process 900 t per day, more than the 500 t they use for cooling each day.

It would be far more worrying if problems like these happened once they're using the system for processing highly contaminated water and not during trial runs with plain water or lowly contaminated water.

It's far preferable to encounter and solve the problems now rather than later.

This testing is what a shakedown is there for. The only problem is the overwhelming time pressure.
Unfortunately, the Kurion cesium removal is in a sequential arrangement with the other modules, so it is now gating the throughput until the source of this blockage is found.
Hopefully, the testing is just done with water, so the testing can be done freely, without needing to take extra precautions because of radioactivity.

 

[causeceleb]

Unfortunately, the indicated temperature will be effected by the emissivity coefficient of each surface being measured. For example, if our iron bar in your example is polished on one end and rough finished on the other, there will be a huge difference in the IR emission detected. The emissivity coefficient will vary with both the surface finish and the type of material.

it seems to me that you all are talking about ancient history.
aren't we presently dealing with 3 or more distinct total meltdowns?
and doesn't "meltdown" imply "corium"? and doesn't "corium" imply
temperatures in excess of 3000 degrees Fahrenheit?

if this is the case, then how do any of these current fact bare on the state of the present
situation at Fukushima?

 

[Bioengineer01]

I would appreciate it if you could provide examples or references to that common regulatory practice. Maybe I shouldn't be so hard on Japanese regulators, if that is happening here.

I am afraid that my experiences are from the health care industry (21+ years) and are mostly derived from conversations with actual regulators in charge of writing the stuff. And even in one instance from a presentation given to engineers by an ex director of the FDA where he told us that directly. I don't have the time to go search the nuclear regulations for examples. Also in the USA things may be different between medical devices and nuclear in terms of safety regulators behavior. I saw an example that suggests that nuclear in the usa may be different and much more lenient to industry. That is Emergency plans. The law stated in the code of federal regulations, nuclear is part 10; 10 CFR, part 100 (reactor site criteria, chapter 11 emergency planning). The law establishes 1 criteria, nobody in the population should get more than 25 REM of radiation during the course of a nuclear accident. The law doesn't specify the distance, just that the population needs to stay below 25 REM exposure no matter what. Now the regulators that need to implement the law make assumptions that are very lenient for the worst case, they assume to create the NEWREGs (0654 and 0396) that the worst case is 1% of fuel failure, that 95% of that failed fuel gets stuck in the containment, and that the 5% of 1% (0.05%) is released at 1/2% per day to come up with the emergency evacuation zone. Also the emergency plans make a lot of assumptions about the status of the infrastructure and the behavior of the people involved that are not realistic. This ends up with small evacuation zones (10 miles), relatively simple and achievable plans (in paper) that are very favorable to industry. In medical devices, it is quite different, if there is a significant problem, no matter what industry did, the regulations are such that the FDA will always be able to find something you did wrong.

 

[tsutsuji]

It sounds like they have problems with one of 4 units, each of which is designed to handle 300 t per day. If the other three units work as expected they would still be able to process 900 t per day, more than the 500 t they use for cooling each day.

That unit has been tested again today after checking the valves and the water flow was found to be OK, although the exact cause of the problem is not clear. The test with low level contaminated water starts tomorrow. The initial plan is to run the test for 5 days but that could be shortened : http://www.asahi.com/national/update/0612/TKY201106120177.html

Taking into account additional contaminated water storing capacity secured at the unit 1 condenser or at the high temperature incinerator 2nd basement, the expected date of overflow is postponed to June 27th : http://www.nikkei.com/news/headline...19595E3E0E2E2E38DE3E0E2E4E0E2E3E3E2E2E2E2E2E2

 

[clancy688]

if this is the case, then how do any of these current fact bare on the state of the present
situation at Fukushima?

They're talking about the IR image of Unit 4. The meltdowns occurred in Units 1-3. Probably no meltdown in Unit 4.

 

[Bioengineer01]

Nothing to be overly pessimistic about neither. Looking at the sheer size and complexity of the facility (and the incredibly fast construction time which is already a feat in itself), this is no surprise that they run into some problems.

We can only wait and hope that the whole thing will work as advertised and, once started, won't run into serious issues.

After somebody told me that the news I posted was NO NEWS... I only wanted to reflect on the status of the project... My project experience also suggests that they were going to run into "expected" problems... but what we don't know is how the schedule was constructed, was June 15th the best date, the expected date or worst case... Now we know it was not the worst case... I am willing to bet based on running this type of level of complexity projects that the expected date is somewhere after end of June and June 15th was shortest possible date if NOTHING went wrong (something that only a PR or an administrator could think). Don't want to speculate on worst date...

 

[Bioengineer01]

They're talking about the IR image of Unit 4. The meltdowns occurred in Units 1-3. Probably no meltdown in Unit 4.

This discussion reflects on the poor credibility record laid out by TEPCO, any apparent inconsistency needs to be explored to the limit to be sure we understand what is happening and the thermal images of Unit 4 have shown the inconsistency of higher temperatures at the site of the RPV/PCV for a long time. I could point to lots of conspiracy theories created around that inconsistency. The witness testimonies of the presence of the JDF at the time of the explosion of Unit 3, aired on live TV and then captured by very few written news sources, and most later taken down from the net, do not help either.

 

[zapperzero]

the NEWREGs (0654 and 0396)

They are called NUREGs and they do recommend planning for the consideration of a 10-mile zone for immediate evacuation or sheltering in place, regardless of accident size.

http://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/sr0396/sr0396.pdf
http://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/sr0654/r1/sr0654r1.pdf

0396 specifically states:
"The task force does not recommend that massive emergency preparedness programs be established around all nuclear power plants" (underlining is not mine, it appears in the original text). while 0654 revision 1 contains such delightful examples of wishful thinking as
"the size of the ingested exposure EPZ (about 50 miles in radius) was selected because [..blah blah other reasons here..] there may be conversion of atmospheric iodine [...] to chemical forms that do not readily enter the ingestion pathway"

I will note in passing that both NUREGs are in PDF format, but the PDFs do not contain text, but rather scans of typewritten pages. Of course, such a format is not searchable or indexable. I do not understand why this is so.

EDIT: NUREG 0396 is dated 1978. It's quite surprising to me that it has not been amended in any way, after TMI and Chernobyl. I suppose nothing was learned since?

 

[Jorge Stolfi]

Unfortunately, the indicated temperature will be effected by the emissivity coefficient of each surface being measured.

Does anyone know how the temperatures are measurd by the IR imaging equipment?

AFAIK, astronomers measure the intensity of the IR radiation at several wavelengths, and infer the temperature of the source from the ratios between those intensities. That allows them to compensate in part for emissivity coefficients, unknown distances and absorption by any cooler stuff that may lie along the line of sight. It is like a blacksmith evaluating the temperature of a hot iron by its color, which (unlike the brightness) does not depend on the status of the surface. But astronomers usually obtain this data by taking separate monochromatic images through different filters.

Are the temperatures in the Fukushima IR images measured by the same principle? I do not know whether there are IR cameras that can take snapshots in two or more wavelengths at the same time (i.e. "color IR" pictures; and the multiple-filter method seems rather awkward for military use.

Could it be that they are using monochromatic IR images, and inferring the temperature from the intensity alone? That method should work in known environments (such as for monitoring equipment in a plant). It may also work for military analysis of aerial images, where it may be enough to detect warm spots, with only crude estimates of temperature based on distance and ambient temperature. But I can imagine many complications in the case of the Fukushima Daiichi images, including the presence of debris over the pools and condensed (hence cooler) "steam" clouds above the reactors. In any case it seems to me that a temperature scale based on IR brightness alone would be an "educated guess" at best.

 

[Jorge Stolfi]

I will note in passing that both NUREGs are in PDF format, but the PDFs do not contain text, but rather scans of typewritten pages. Of course, such a format is not searchable or indexable. I do not understand why this is so.

Because Postscript was developed only in the late 1980s, and became reasonably popular only in the next decade. Therefore much of the literature before 1990 is available only on paper or microfilm, and the only reliable way to digitalize it (including diagrams and subtle typographic details) is in the raster image format.

One can run an OCR on those images to obtain also a text repesentation that can be searched and copy-pasted, and include it in the PDF file so as to imitate some the functionality of true PDF format. Technical journals have been doing that with their past issues, and Google was doing the same with old books and magazines from libraries. However the OCR output one gets from old print is not so good, and it takes a lot of manual editing to make it usable for searching.

 

[westfield]

Does anyone know how the temperatures are measurd by the IR imaging equipment?

snip .

For those interested, "NEC \ AVIO" cameras are mentioned in the documents, http://www.nec-avio.co.jp/en/products/ir-thermo/"

 

[westfield]

This story is of precisely zero importance.

snip .

Well this is a physics forum not a psychics forum, we still need to hear about it ;)

 

[jim hardy]

The few IR cameras i ever used had a knob on the side labelled "Emissivity" and when you turned it it changed all the reported temperatures.

So I take IR pictures to be a relative measurement showing approximate distribution of heat, not like a thermometer showing true temperature.

Emissivity is something's affiinity for giving off heat via radiation.

That's why a Franklin stove is flat black(higly emissive) and the inside of a Thermos bottle is silvery white so as to not radiate well.
Everything else is in-between.

I have no idea whether liquid water is transparent or opaque to IR, or in between. So those IR cameras could be showing me either the surface of the water or the bottom of the pool, and i'd defer that pronouncement to somebody who does know .

 

[radio_guy]

I still hold the view that using IR/thermal imagery from a closer distance could give them a better picture of where the core debris is located and they could possibly modify the cooling plan based on it's location. maybe an on site set of images from different angles could give enough perspective into what they are really dealing with, it could also be used to pinpoint leaks and help them with at least having a better idea of the current failure modes.

I don't know about any of you, but if I was the plant operator and was applying cooling and making plans on a best educated guess I would be looking for ANY other form of information other than "this seems to be working for now, let's keep doing it" right now from what I am reading is they are working on a system of action-reaction, change something and see what happens then base changes on that.

 

[ceebs]

Here it is: http://www.tepco.co.jp/cc/press/betu11_j/images/110603a.pdf (page 8)

Out of the 720k TBq, 140 TBq account for C137. A little comparison: The Chernobyl core had a total C137 inventory of 280 TBq (of which only 40% escaped).

So is that saying there's more C137 in that trench, than escaped from Chernobyl?

 

[Atomfritz]

I will note in passing that both NUREGs are in PDF format, but the PDFs do not contain text, but rather scans of typewritten pages. Of course, such a format is not searchable or indexable. I do not understand why this is so.

EDIT: NUREG 0396 is dated 1978. It's quite surprising to me that it has not been amended in any way, after TMI and Chernobyl. I suppose nothing was learned since?

One can run an OCR on those images to obtain also a text repesentation that can be searched and copy-pasted, and include it in the PDF file so as to imitate some the functionality of true PDF format. Technical journals have been doing that with their past issues, and Google was doing the same with old books and magazines from libraries. However the OCR output one gets from old print is not so good, and it takes a lot of manual editing to make it usable for searching.

Some things aren't being OCRed in the public versions.
Too many curious people could find things that some do not want to be public.
Just a quite obvious example:

More than warming not heeded, the explanation is different, it is a common regulatory practice, even in the USA of not defining in hard numbers the limits when those are known to be too high due to cost considerations and leaving the decision making to Industry, fully knowing that they will be forced to compromise. The problem with NP is that the final liability is taken by taxpayers, differently from other industries.

I would appreciate it if you could provide examples or references to that common regulatory practice. Maybe I shouldn't be so hard on Japanese regulators, if that is happening here.

I don't have the time to go search the nuclear regulations for examples.

Just let's take a document written on a word processor as example.
A document that's origin is digital (not typewriter!) and is probably stored fulltext in the NRC databases.

But for the public it has been printed out and then scanned as image to be published as non-indexable (and so non-googleable!) PDF.

This document is called "NUREG CR/5969" (thanks to the guy who recently posted a http://www.ornl.gov/info/reports/1992/3445603689514.pdf" to!).
It deals with strategies to avoid/mitigate severe BWR accidents.
In the section discussing recriticality issues and weaknesses of some boron injection systems you read on page 59, regarding suggested safety upgrades:
"However, this would require significant modifications to existing equipment and this would be beyond the scope of the present study."

In clear text this means:
"It would cost the reactor industry money to make sure that this recriticality issue is resolved.
So this is a no-no, and we won't discuss this issue."

Just one of a plethora of examples.

 

[jmelson]

d

I have no idea whether liquid water is transparent or opaque to IR, or in between. So those IR cameras could be showing me either the surface of the water or the bottom of the pool, and i'd defer that pronouncement to somebody who does know .

A few hundred feet of moist air can absorb a lot of thermal IR. Pools of liquid water will certainly
absorb ALL thermal IR, so you can only read the temperature of the surface of the pool.

Jon

 

[StrangeBeauty]

Via ex-skf:
http://ex-skf.blogspot.com/2011/06/radioactive-strontium-from-groundwater.html

Strontium found at sub-drains of #1 and #2 and at other locations:
http://www.tepco.co.jp/en/press/corp-com/release/betu11_e/images/110612e10.pdf

It is the first time that radioactive strontium was detected from the groundwater at Fukushima I Nuclear Power Plant.
...
In addition, strontium-89 and -90 were detected from the seawater samples taken on May 16 at 3 locations: at the water intake canals for the Reactors 1 through 4, inside the silt screen for the Reactor 2, and inside the silt screen for the Reactor 3.
...
The amount of strontium-90 in the seawater inside the silt fence for the Reactor 3 was 240 times as high as the legal limit allowed for nuclear plant operations.

Also, it appears to me that none of these graphs portray the expected half life curve of I-131 v Cs isotopes...?
http://www.tepco.co.jp/en/press/corp-com/release/betu11_e/images/110612e11.pdf
Please correct me if I'm wrong (that is, does the first graph conform to expectation? Many others appear not to be following the expectation of I-131 decrease given that the I-131 is not being replenished). Thanks.

 

[Bandit127]

Bioengineer has already alluded to the regulations that the health care industry is subjected to. Obligations exist for the control of documents that make the paper versions simpler to use.

For those reasons, the 'paperless office' is still a long way from reality in health and all the businesses I have seen still use printed versions of some critical documents as the master versions. The better companies are further along the road to an electronic system, but none are paper free. (E.g. validation documents - even the best still have these on paper).

It could well be that simple. Their system might hold the paper copy as the master, in which case it would be entirely natural to scan a printed copy for distribution. As I would if I wanted to email a validation report.

I have no understanding of the obligations for document control in the nuclear industry, but where I work the distribution of a scanned to PDF document is the preferred method.

 

[MadderDoc]

The few IR cameras i ever used had a knob on the side labelled "Emissivity" and when you turned it it changed all the reported temperatures.

So I take IR pictures to be a relative measurement showing approximate distribution of heat, not like a thermometer showing true temperature.

Emissivity is something's affiinity for giving off heat via radiation.

That's why a Franklin stove is flat black(higly emissive) and the inside of a Thermos bottle is silvery white so as to not radiate well.
Everything else is in-between.

That is of course true, and it is unavoidably part of the game when interpreting thermographic images. In simple terms -- the lower the emissivity of the material in the surface layer of an object, the more will the temperature readings from thermography be liable to underestimate the true temperature of the object. One should not overemphasize this as a problem in the case of Fukushima however, since most of what we are looking at there will have a relatively high emissivity, meaning the temperature differences we see in the thermographs are likely to be an effect of true temperature differences.

I have no idea whether liquid water is transparent or opaque to IR, or in between. So those IR cameras could be showing me either the surface of the water or the bottom of the pool, and i'd defer that pronouncement to somebody who does know .

Water is opaque in large portions of the IR spectrum, so that in effect while looking at water in IR you will see only the radiation coming off the surface layer. Water also has a high emissivity, so the temperature reading using thermography will with good approximation be the true temperature of the surface layer of the water.

 

[Joe Neubarth]

My nuclear power theory was all taught to me in naval nuclear power school half a century ago so I am asking a question that we were not taught an answer to. We were taught to operate plants, not deal with theory of glowing radioactive blobs.

At Fukushima we have heard talk of recriticality.

How is it possible for those melted cores to go critical again when the normal operating process was for neutrons from fission to slow down in water and thus interact with Uranium 235. If the water ain't there, are the neutrons becoming thermal (lower energy state) by bouncing off of the entrained impurities in the melted blob? Does the size of the blob provide for reflection of energy (Neutrons) back towards the center?

We know that Reactor One is melted down, yet it sure looks like it is "breathing" (as some call it) with increases in temperature coming in long waves that would seem to indicate increases in criticality and then decreases.

NucEng, can you or somebody else with experience answer this?

Has any institution done any research in Blob criticality theory?

 

[seeyouaunty]

How is it possible for those melted cores to go critical again when the normal operating process was for neutrons from fission to slow down in water and thus interact with Uranium 235. If the water ain't there, are the neutrons becoming thermal (lower energy state) by bouncing off of the entrained impurities in the melted blob?

Earlier in the thread someone mentioned that the moderator (water) doesn't need to get within the corium to slow neutrons, surrounding the corium in water would still allow neutron moderation to occur.

 

[clancy688]

Earlier in the thread someone mentioned that the moderator (water) doesn't need to get within the corium to slow neutrons, surrounding the corium in water would still allow neutron moderation to occur.

How would that be possible? The neutron comes out of the corium at one helluva speed, slows down in the water, makes an U-turn and fissures an uranium atom?

 

[Joe Neubarth]

Earlier in the thread someone mentioned that the moderator (water) doesn't need to get within the corium to slow neutrons, surrounding the corium in water would still allow neutron moderation to occur.

But if they are going out of the core to hit the water, who sends the neutrons back into the core?

I remember in Nuc Pwr Sch they did tell us that the reactor vessel does act to redirect the neutrons back in the direction of the core if they try to escape. Gotta herd them little doggies in the right direction or they might get lost.

But if the Blob is out of the reactor, the neutrons would be free to wander in the cement and mortar around them until the blob hits bedrock.

Hummmmm? Blob Criticality Theory. Somebody has to be writing a thesis on this for a doctorate. Right?

"It's Alive! It's Alive!"

 

[tyroman]

This discussion reflects on the poor credibility record laid out by TEPCO, any apparent inconsistency needs to be explored to the limit to be sure we understand what is happening and the thermal images of Unit 4 have shown the inconsistency of higher temperatures at the site of the RPV/PCV for a long time. I could point to lots of conspiracy theories created around that inconsistency. The witness testimonies of the presence of the JDF at the time of the explosion of Unit 3, aired on live TV and then captured by very few written news sources, and most later taken down from the net, do not help either.

@ Bioengineer01


Fine, go and do the exploration... I would suggest you first explore the possibility that there is NO inconsistency in the thermal signatures of the SFP versus the visible surface of water in the reactor vessel. This exploration should include determination of the temperature (ambient) of the lake water which TEPCO has been pumping into the SFP, how much of this water has been pumped into the SFP (and of that total, how much accidentally got into the reactor vessel), the total decay heat being generated by fuel in the SFP, the level of residual radiation in components contained in the reactor vessel, the thermal output of such residual radiation, a calculation of the insulating characteristics of walls and floor of the SFP, a similar calculation for the reactor vessel, a determination of the heat lost from the water surfaces of the SFP and the reactor vessel... and so on.

The thermo calculations would be rather straight forward. Some assumptions would have to be made in the case of data which is not available, but a ballpark figure for what the relative thermal signatures should be is achievable in my opinion. These theoretical results would of course need to be adjusted for variables influencing the IR signals received by the satellite... such as the amount of steam at one location versus another. The images which have raised the spectre of conspiracy in your mind, may in fact be completely consistent with the real world.

My point is that (in trying to build a conspiracy theory around an "inconsistency" which probably does not exist) you are comparing apples and oranges. The two pools of water exist in hugely different environments and have significantly different histories.

.

 

[swl]

it seems to me that you all are talking about ancient history.
aren't we presently dealing with 3 or more distinct total meltdowns?
and doesn't "meltdown" imply "corium"? and doesn't "corium" imply
temperatures in excess of 3000 degrees Fahrenheit?

if this is the case, then how do any of these current fact bare on the state of the present
situation at Fukushima?

It is my understanding that the only fuel in unit 4 resides within the SFP. It's location within the SPF in no way implies that it is safe there and insignificant.

 

[Jorge Stolfi]

Does the size of the blob provide for reflection of energy (Neutrons) back towards the center?

How would that be possible? The neutron comes out of the corium at one helluva speed, slows down in the water, makes an U-turn and fissures an uranium atom?

My understanding is based on an analogy with optics. A white material such as milk, chalk, white paint, or clouds looks white because all the light that falls on it is scattered back to the open space on the illuminated side, in a random direction --- instead of being absorbed or transmitted to the other side. However, at a microscopic scale, most white materials consist of transparent particles in a transparent medium with a different index of refraction. So, a large part of the incident light is not scattered off the material 's surface; instead it enters the material, and is then refracted or reflected by its particles many times, random-walk style. The internal scattering has no preferred direction; but if there is no absorption and the material is thick enough, most of the photons will eventually come out of it *towards the same side they entered*. In other words, their U-turn is not deterministic but only a statistical event --- due to the fact that there is no scattering on one side of the surface.

By the same token, if one were to suspend a light bulb by a thin wire in the middle a thick fog bank, much of the light emitted by it will eventually come back to it --- simply because the wandering photons cannot end their trip in any other way. This is more obvious if the lamp is suspended in milk; but it should seem reasonable also for fog, when one considers that a thick enough cloud bank at noon will look black from below.

The same should happen to neutrons that escape a fuel mass completely surrounded by a large body of water, or any other moderating material: as long as we can ignore absorption and spontaneous decay in the water, most of them will eventually come back to the fuel, with thermal energies.

However, if there is some absorption in the surrounding material, not all neutrons will come back. This is analog to what happens to photons in materials like smoke or cappuccino, that contain light-absorbing substances together with light-scattering grains: the material will look gray or colored, instead of white. Also, if there are other absorbing objects (such as a concrete floor) immersed in the same medium and not far from the fuel, part of the neutrons will end up there before coming back to the source.

In the case of Fukushima, I believe that they are using borated water to cool the corium, instead of plain water; which in the optical analogy should be equivalent of suspending a lamp in china ink instead of milk, or in tire smoke instead of fog. However the uranium in the molten mass may be covered by some other material (iron and zirconium oxides?) whose "neutron color" I cannot guess.

As I said, my understanding is all based on this optical analogy. It may not apply at all to the situation.