Japan Earthquake: Nuclear Plants at Fukushima Daiichi

[Ms Music]

People in Seattle got on average 5 hot particles per day for the month of April, 2011.
(acc. to Arnie Gundersen, Fairewinds Associates)
Question: What's the known effect of around 5 hot particles a day? The most I can find is the following, indicating a "big" risk of leukemias and cancer - from a study done in the 70s:

http://www.nuc.berkeley.edu/node/4459

Or, anyone with nuclear plant experience know anything about hot particles and what their effects might be?

Many thanks.

This is what has been recorded in Seattle since March.

http://www.doh.wa.gov/Topics/japan/monitor-history.htm


I personally am more concerned about the two flights I took, and the full body scans to make sure I wasn't hiding anything in my bra.

 

[Quim]

This is what has been recorded in Seattle since March.

I agree with you about the distant contamination via the jet stream etc.

But lately I am becoming more and more concerned about how much junk is being dumped in the Pacific.

The major portion of the "local" contaminants went out over the ocean.

Then there is the water release, and I don't trust TEPCO to not be dumping radioactive liquids in the ocean on an ongoing basis.

From what we have learned about the geology of the site recently, all the radioactive material which finds its way into the groundwater or runoff water finds its way into the Pacific.

I am growing uncomfortable with this idea. I think there should be somebody there monitoring TEPCO.

 

[swl]

I guess I have to say it again. If you contain the buildings, it allows you to filter the venting instead of letting it drift down wind. That is justification enough. As to hiding what is going on iside a containment "tent," don't you think by now that we already know something bad happened there? If offsite dose rates decline or increase isn't that good enough to tell whether their efforts are working?

All excellent points. Could you please post a link describing the HVAC system they'll be using? I'm curious how they expect to control RH/condensation and temperature while preventing escape of contamination.

 

[Ms Music]

But lately I am becoming more and more concerned about how much junk is being dumped in the Pacific.

I think I read in the last day or two that they have detected trace amounts of radiation in whales off Japan?

 

[MadderDoc]

C137 release data for Units 1, 2 and 3:

Unit 1: 5.9E14
Unit 2: 1.4E16
Unit 3: 7.1E14
Official report, attachement IV-2, page 7.

The contamination outside the plant is nearly entirely the fault of Unit 2. Unit 1 and 3 may have been eyecandy, but Unit 2 is the real headache.

Those figures appear to be the results of a modelling attempt based on a selected subset of possible scenarios, and not based on actual release measurements. You cannot base such a strong statement on that.

 

[Jorge Stolfi]

It seems so obvious that there were multiple explosions. Just look at this http://www.tepco.co.jp/en/news/110311/images/110611_05.jpg" from floor 4 of RB#4. (please open in a separate tab, big size!) Look at the 480V distribution you see in the lower right. [...] Can be there any doubt that there was a chain-reaction of explosions?

Perhaps. But (as old jim observed) the hydrogen probably had enough time to penetrate the electrical boxes. Even if the H2 inside the box and that in the room exploded at the same time, I would expect the box walls and doors to be pushed outwards.

Also, could there have been pressure pulses traveling down the cable conduits from another floor, ahead or behind the main explosion shockwave in the room? The route of the latter (say, from the service floor and through the fuel elevator well) may have been longer or shorter than the route through the electrical pipes.

 

[MadderDoc]

I have posted my theory about the Unit 3 explosion(s) in the "What caused the big explosion at fuku reactor three" thread here;
https://www.physicsforums.com/showpost.php?p=3358847&postcount=104

I was prompted to post this theory by the TEPCO Unit 3 video recently linked on this forum.
Video is at:
http://www.tepco.co.jp/en/news/110311/images/110615_01.zip

Please review and give your opinion at the other thread.

.

The other thread has been locked by a moderator with reference to your posting, so I won't be able to comment any further there until that thread has eventually been unlocked. (Taking it to this thread would seem not right, unless the moderator of this thread specifically would allow it.)

 

[razzz]

Woods Hole https://www.whoi.edu/page.do?pid=68736" link is very good reading (all days) albeit premature on results.

 

[Jorge Stolfi]

My little theory about Unit 2: Molten corium in the primary containment failed the walls to the torus room and/or second floor in the eastern side of the reactor building which resulted in a sudden depressurization and a hydrogen explosion bursting into the turbine building, thus connecting the primary containment of Unit 2 directly with the outside.

It seems unlikely to me that the corium could reach the torus directly. There is a partial concrete barrier inside the drywell, below the RPV (the hollow RPV pedestal), that should constrain the corium to fall at the bottom of the drywell, even if it squirted sideways out from the RPV.

Moreover, the vent discharge galleries that connect the drywell to the torus start a foot or two above the drywell floor (in the #1 blueprints at least, may not be true for #2--#4); so it should take a lot of corium to overflow that barrier.

My guess is that the damage was caused simply by a pressure surge in the primary containment (drywell+galleries+torus, which are basically a single cavity), above the maximum design pressure of ~600 kPa (~6 bar); and the weakest spot (the one that burst) happened to be somewhere in its lower portion.

 

[Jorge Stolfi]

The basement rock of granite type material starts about 800 meters below the power plant and is overlain by the Tomioka Formation of Late Miocene/Early Pliocene sedimentary rocks upon which the the plant is built. ... Now will somebody please tell me that these do not represent any kind of earthquake hazard.

My impression from this thread is this:

* The rock below the reactors is solid enough to support them during any earthquake, even the 3/11 one. Given their "floating box", bottom-heavy construction, I cannot see any significant risk of them collapsing, cracking, or even tilting just because that rock is "not strong enough".

* There is some risk of the reactor buildings cracking if the underlying rock itself cracks and the two sides move relative to each other, as in a fault. Photos of such "mini-faults" on the ground near the reactors have been posted here, so perhaps this scenario has already happened. However, the floating box construction seems to give some protection against that possibility. Also, such fault-like cracks presumably start deep underground, so they should occur with the same relative displacement and same probability, no matter what the rock type.

* The type of rock matters for issues related to underground water (such as leaks into or out of the building). However, the underground water flow at Fukushima Daiichi seems to be slow and directed towards the ocean. So any leakage through that route should not spread inland, and should be small compared to the oceanic contamination that has already occurred.

* The type of rock has a major influence on the amplitude (and hence acceleration) of the ground motion during an earthquake. However there is no need to speculate about this point, since TEPCo very early on published the maximum ground acceleration registered in each building during the 3/11 earthquake (and, IIRC, some of those numbers were well above the values assumed in the plant's design).

Does this make sense?

 

[Quim]

It seems unlikely to me that the corium could reach the torus directly.

I'll chime in with agreement on that. After all that's why GE used a torus for the blow down pool.

My guess is that the damage was caused simply by a pressure surge in the primary containment

Agree here too, this is one of the few things we know about unit #2, although there was a hydrogen explosion, the containment held but developed some leaks apparently.

The odd thing about unit two is that it somehow found enough oxygen inside the primary containment to create a significant explosion.

I have looked into what I believe the environment in #3 was like just before it blew and from what I surmise, it was Hydrogen rich and steam laden but it had to have been very very short of oxygen or it also would have gone off inside containment, (this is probably similar in unit one, but I haven't looked into the data there in detail yet.)

Where did unit #2 get the oxygen from?

 

[Jorge Stolfi]

He probably wants to make sure that cigarettes won't be the thing which kills him. (We need to check on him over the next fifty years. If he dies of cancer, I'll submit the request for a Darwin Award. :D )

Sorry if this straddles off-topic, but... I have seen several reports that many of those people are not professional nuclear workers, but instead people from all walks of life who have been trained only to do specific tasks. I would not be surprised if it turns out that they do not know what a sievert is, or the difference between alpha, beta, and gamma, or that the smoke that he inhales from the cigarette is just as dangerous as the surrounding air.

So that poor worker's lapse may not be entirely his fault.

 

[Most Curious]

Hopefully someone here with first hand knowledge of similar plants will confirm or disprove my thoughts.

There are many piping and electrical penetrations into the drywell and probably some into the torus as well. I am assuming those penetrations are sealed with a flexable material that has a maximum temperature limit, hence the maximum temp. rating of containment. May we assume that with no power, therefore no cooling of containment with a VERY hot reactor inside that most or all of those penetration seals were cooked early on? If so, movement of air into and out of containment would occur by convection if pressure reached atmospheric..

 

[LabratSR]

Hopefully someone here with first hand knowledge of similar plants will confirm or disprove my thoughts.

There are many piping and electrical penetrations into the drywell and probably some into the torus as well. I am assuming those penetrations are sealed with a flexable material that has a maximum temperature limit, hence the maximum temp. rating of containment. May we assume that with no power, therefore no cooling of containment with a VERY hot reactor inside that most or all of those penetration seals were cooked early on? If so, movement of air into and out of containment would occur by convection if pressure reached atmospheric..

Enjoy. CEPA = Containment Electrical Penetration Assembly

http://www.osti.gov/bridge/servlets/purl/5338909-FPWlHy/5338909.pdf

 

[Quim]

My impression from this thread is this:

* The rock below the reactors is solid enough to support them during any earthquake, even the 3/11 one. Given their "floating box", bottom-heavy construction, I cannot see any significant risk of them collapsing, cracking, or even tilting just because that rock is "not strong enough".

* There is some risk of the reactor buildings cracking if the underlying rock itself cracks and the two sides move relative to each other, as in a fault. Photos of such "mini-faults" on the ground near the reactors have been posted here, so perhaps this scenario has already happened. However, the floating box construction seems to give some protection against that possibility. Also, such fault-like cracks presumably start deep underground, so they should occur with the same relative displacement and same probability, no matter what the rock type.

* The type of rock matters for issues related to underground water (such as leaks into or out of the building). However, the underground water flow at Fukushima Daiichi seems to be slow and directed towards the ocean. So any leakage through that route should not spread inland, and should be small compared to the oceanic contamination that has already occurred.

* The type of rock has a major influence on the amplitude (and hence acceleration) of the ground motion during an earthquake. However there is no need to speculate about this point, since TEPCo very early on published the maximum ground acceleration registered in each building during the 3/11 earthquake (and, IIRC, some of those numbers were well above the values assumed in the plant's design).

Does this make sense?

Yes it does, and there is also the porosity issue with the rock under Fukushima.

I've been interested in the geology of the site too, but not because I think the structures are in a mechanically unstable condition, I don't see any problems there. It would have been nice to have the foundation linked with solid granite - that would minimize shake for several reasons. But I can't fault the placement of the site on those grounds.

Where my interests lie are in the drainage and underground water paths. At first I was concerned that this accident could contaminate the groundwater over a huge area of the island with tragic result, but it turns out that that is not the case.

The mountains just to the west have an established runoff pattern which is not the worst of all possible alternatives. This rock has a substantial degree of permeability. The water flow is in a steady and slow passage to the ocean through the rock underlying Fukushima.

So the Japanese groundwater is safe.

 

[jim hardy]

Tyrol and Madderdoc

if you've not already read it - this may help with your hydrogen inquiry.

http://www.iaea.org/inis/collection/NCLCollectionStore/_Public/25/028/25028788.pdf

Hydrogen Problems Related to Reactor Accidents
A.Bujor

you may not need to condense the steam, see around p 12.

old jim

 

[a|F]

My little theory about Unit 2: Molten corium in the primary containment failed the walls to the torus room and/or second floor in the eastern side of the reactor building which resulted in a sudden depressurization and a hydrogen explosion bursting into the turbine building, thus connecting the primary containment of Unit 2 directly with the outside.

You lack understanding of the interrelationship between the reactor, drywell, and torus. As far as #2's explosion, I'll stick with this, sourced multiple times in this thread:

"it is realistically assumed that without operator action, reactor vessel pressure control over the long term would be by repeated cycling of the same relief valve.* Because of the high steam mass flux into the suppression pool bay in which the discharging T-quencher is located, significant thermal stratification would be expected. MARCH computations show that the difference between the local and average suppression pool temperatures can be estimated to increase from about 5°C at the beginning of the transient to about 40°C 100 minutes later. This means that the suppression pool would lose its condensation effectiveness; the resulting pressure loads from the SRV discharge of steam and noncondensibles would rapidly increase, leading to a possible rupture of the wetwell which could occur before the overtemperature-induced failure of the drywell."

http://www.ornl.gov/info/reports/1981/3445600211884.pdf

 

[clancy688]

It seems unlikely to me that the corium could reach the torus directly. There is a partial concrete barrier inside the drywell, below the RPV (the hollow RPV pedestal), that should constrain the corium to fall at the bottom of the drywell, even if it squirted sideways out from the RPV.

I posted an old american study of Mark-I containments in my thread which concludes that exactly that may be on of the highest likely happenings in case of meltdowns.

Those figures appear to be the results of a modelling attempt based on a selected subset of possible scenarios, and not based on actual release measurements. You cannot base such a strong statement on that.

Still it isn't the first time that something's claiming that Unit 2 is entirely at fault. I remember such rumours spreading out since March.


And why don't why outsource everything regarding Unit 2 in the news thread? Whether my train of thought may be faulty or not, the thread is there now, let's use it.

 

[Orcas George]

It would be nice to see some solutions offered to as what would work and how to approch it. There maybe a thread that is all ready disscusing this, if so could you please direct me. I do not know how to navigate this forum, usually just read.

If the idea is to contain it, I would first think of containing a smaller area. Every foot that you add to a structure increases its complexity and I think there is some sort of exponential law on sail area. Wrapping a four(?) story building in a typhoon zone is a very ambitious project, making that wrapping air-tight is unheard of, and constructing it all by remote control is a fantasy in my opinion. I can imagine creating a lightweight shell out of sections made of modern FRP composites (kevlar, carbon) but I cannot imagine how to make the sections airtight.

The right answer is to get the hot spots cooled below boiling point using a closed loop system. I would have imagined that they would use the very large portable industrial cooling systems (container sized and larger) that you can rent to respond to industrial emergencies, but there must be some reason that makes that unworkable. It seems like you could drop a bunch of loops with good old-fashioned freon (and then out to an air-cooled refrigerator). Who knows, perhaps freon does something nasty when it is irradiated.

 

[tyroman]

Tyrol and Madderdoc

if you've not already read it - this may help with your hydrogen inquiry.

http://www.iaea.org/inis/collection/NCLCollectionStore/_Public/25/028/25028788.pdf



you may not need to condense the steam, see around p 12.

old jim

Thanks old jim,

Yes I am familiar with the "Flammability and detonation limits" diagram on page 12 of your link.

I first saw it on page 7 of 42 of a document linked at Homer Simpson's post:
https://www.physicsforums.com/showpost.php?p=3231637&postcount=2996

The document:
http://canteach.candu.org/library/20044507.pdf
contained the same "flammability limits curve" I refer to in the "theory" I have been chastized for posting in the Joe Neubarth thread which was established to air "theories" and speculation about "What caused the big explosion at fuku reactor three".

Until that issue is settled, I will defer further comment...

.

 

[NUCENG]

Good explanation, thanks!

One thing I don't understand about theormocouples is why the bimetal wires are usually extended all the way to the voltmeter. Why couldn't they be extended only to some cooler place nearby (such as just outside the concrete enclosure), and then have the signal be carried by copper wires to the meter? That would result in lower resistance for the signal and reduced risk of electrochemical effects along the way.

Or is that in fact how it is done?


Thanks for the compliments and for the thorough analysis. As for them being "open", there is a continuum between having a small leak and being wide open, so it may not be a simple yes/no question. Also, for a small leak, the degree of opening may be sensitive to pressure, temperature, flooding, clogging, etc., and so may vary erratically with time.


Thanks! I think I saw mention of it in this forum, but hadn't the time to check it out then.

Thanks again. I am tempted to include those readings in my plots too, but first one question: do they reflect the conditions inside the reactor, or only of the external contamination? In other words, are those gammas and neutrons mostly created by fission and decay inside the reactor's concrete enclosure? If so, does the spent fuel in the SFP contribute to those readings?

Thanks, that is important information.

As for the temperature sensors, I have seen several diagrams showing their approximate location on the RPV, drywell and torus; but I still miss the key details. Namely, where precisely are the RPV temperature measured: on the outside surface of the RPV, or embedded into its wall? If the former, woud the reading be affected by the drywell atmosphere or by water leaks above the sensor? How far is the "water nozzle" temperature sensor from the nozzles and their feedpipes? And so on...

These details are important, for example, to analyze the pressure x temperature plots. The red boiling curve in those plots is relevant only if the temperature and pressure are measured at the same spot in the fluid. Barring gauge malfunctions, the pressure must be indeed that of the fluid at the gauge's intake point, which should be valid for the bulk space inside (except for the hydrostatic pressure gradient in the liquid-filled part). On the other hand, if the temperature is measured on the outside of a 15 cm thick wall, or even embedded into it, it will be some value intermediate between the temperatures of the two fluids in immediate contact with the wall. Thus, one can easily have superheated steam inside the RPV with a temperature reading well below the boiling curve, or (less likely) liquid water inside with a temperature reading well above the boiling curve.

If you are still looking for thermocouple locations try this:
http://www.tepco.co.jp/en/nu/fukushima-np/f1/images/measuring_positions-e.pdf

 

[Quim]

If the idea is to contain it, I would first think of containing a smaller area.

I agree completely with you on that.

The right answer is to get the hot spots cooled below boiling point using a closed loop system.

But this is an impossibility. The foundations are cracked and leaking. Radiation laden water is being fed into the porous rock below and will spread, and it will end up in the ocean. It is in our best interests to keep the contaminated area minimal.
IMO Trying to pour water on the corium piles is a fools errand.

Here is the profile of the site that was drawn for us by tonio, who responded to the request for geological information (post #9285.)
[PLAIN]http://img30.imageshack.us/img30/5323/geoprofile.png

We need to keep that plume as small as possible.

The piles are now only generating a megawatt of heat each or less.
Its time to let the piles reach a heat equilibrium with their environment.
TEPCO needs to pour sand (or whatever they think will mix with the corium and dilute it further) in with the water.

Then stop watering it.

IMO

 

[zapperzero]

TEPCO needs to pour sand (or whatever they think will mix with the corium and dilute it further) in with the water.
IMO

Leaving aside the rather obvious problems with pumping sand through already-stressed pipes that were designed for ultra-pure water.

How do you propose the mixing would occur? The likeliest scenario is that the molten corium is now covered in a solid crust, due to having been cooled constantly and being a rather poor conductor of heat in the first place...

 

[~kujala~]

I started a new thread for geological issues:

https://www.physicsforums.com/showthread.php?t=507369

 

[NUCENG]

All excellent points. Could you please post a link describing the HVAC system they'll be using? I'm curious how they expect to control RH/condensation and temperature while preventing escape of contamination.

I just downloaded a bunch of stuff related to the erection of the containments and will see if they discuss the filtration system. In filtration of humid air they would probably use an approach similar to the SBGT system which first passes the stream through a heater that lowers relative humidity, then through a HEPA filter to remove particulates and finally through activated charcoal to remove gaseous and ionic radioactivity.

 

[NUCENG]

I started a new thread for geological issues:

https://www.physicsforums.com/showthread.php?t=507369

Just a suggestion cross post a ling to the new thread on the Earth Forum. They have been very responsive to my question about mudstone.

 

[SteveElbows]

I have little inclination to engage in any science based discussions of the Fukushima events external to this thread.

Oh, I am just about to respond to things you and Quim said here about reactor 2, but I'm doing it in the other thread.

 

[MadderDoc]

Isn't there a doubt that not only two explosions happened?
It seems so obvious that there were multiple explosions.

Just look at this http://www.tepco.co.jp/en/news/110311/images/110611_05.jpg" from floor 4 of RB#4. (please open in a separate tab, big size!)
<..> to me, this picture does not prove much except just that there were multiple explosions...

Can be there any doubt that there was a chain-reaction of explosions?

It stands to reason that a hydrogen explosion in a complex physical system like the Fukushima reactor building no 4 may very well not be uniform throughout the building and all of its crevices. It is a matter of language use whether one will label this natural 'graininess' as multiple explosions. When, in relation to another unit, thoughts have been raised regarding 'two explosions' I think that is referring to something else, namely the possibility of two interacting, but physically different mechanisms of explosive force.

 

[SteveElbows]

Looks like the new plan to provide unit 4 fuel pool cooling involves a pretty basic solution, attaching the end of a hose to some railings on the refuelling bridge.

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_03-e.pdf

Also the levels of Cs-134 and Cs-137 measured at reactor 2 building are not exactly plummeting since they opened the doors and did some filtering & ventilation:

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_01-e.pdf

 

[SteveElbows]

Also the levels of Cs-134 and Cs-137 measured at reactor 2 building are not exactly plummeting since they opened the doors and did some filtering & ventilation:

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_01-e.pdf

Oops I probably shouldn't have mentioned opened doors, since that's the next stage:

http://www3.nhk.or.jp/daily/english/16_23.html

Any ideas how they are able to give such positive numbers in that press article? Am I reading the graph wrong or are they using out of date info?

 

[MadderDoc]

<..>

"Also the levels of Cs-134 and Cs-137 measured at reactor 2 building are not exactly plummeting since they <..> did some filtering & ventilation:

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_01-e.pdf

http://www3.nhk.or.jp/daily/english/16_23.html

Any ideas how they are able to give such positive numbers in that press article? Am I reading the graph wrong or are they using out of date info?

My best idea is that the numbers in the press article might be a -- perhaps somewhat positively skewed -- representation of data as it looked a few days ago, maybe when a Tepco application was made to authorities -- I imagine opening the reactor building to the environment would be one of those things that the utility would need official permission to do.

It seems clear from the Tepco press releases, that the full data set you reference includes the time for initiation of the air purification, quote Tepco,
"-We installed local exhausters and started to operate them at 12:42 pm on
June 11 in order to improve the working environment inside Reactor
Building of Unit 2."
and the full dataset then does not seem to support that there has been a reduction of the Cs isotopes over the period of operation so far, nor that the reduction of I-131 has had the magnitude expressed in the press article.

 

[joewein]

Looks like the new plan to provide unit 4 fuel pool cooling involves a pretty basic solution, attaching the end of a hose to some railings on the refuelling bridge.

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_03-e.pdf

One of the ripped up pipes in this image at unit 4 was originally planned to be used for the closed loop cooling:

[PLAIN]http://img824.imageshack.us/img824/6213/11061104.jpg

Since the hot water is taken from the skimmer tank, the damaged pipe was probably the other end, i.e. the one feeding back into the pool. That's what the above railings attachment is about.

Also the levels of Cs-134 and Cs-137 measured at reactor 2 building are not exactly plummeting since they opened the doors and did some filtering & ventilation:

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_01-e.pdf

This is speculation on my behalf, but perhaps the cesium in the air is steadily being replenished from evaporation of contaminated water.

For example, if steam was bubbling through hot water, or there was water leaking from a pipe above the ground so that evaporation was not happening from a quiet surface of standing water then dissolved solids would always end up as dust in the air.

 

[Jorge Stolfi]

If you are still looking for thermocouple locations try this:
http://www.tepco.co.jp/en/nu/fukushima-np/f1/images/measuring_positions-e.pdf

Thanks!

 

[MadderDoc]

Looks like the new plan to provide unit 4 fuel pool cooling involves a pretty basic solution, attaching the end of a hose to some railings on the refuelling bridge.

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_110616_03-e.pdf
<..>

One could say the concrete pump method used so far is also pretty basic. and then there is the KISS principle. The alternative sfp injection there presented is of course not in principle different than using the concrete pump and it should work just as well. As a practical matter it would mean less obstruction for the ongoing work at the south face of the building to null the need to have the concrete pump around.

As an aside (see attachment) -- there is a square structure indicated in the drawing at what would be the east side of the unit 4 pool. I wondered if there is any significance to that depictured detail, as regards how the unit 4 pool set up is or is to be actually configurated. (The sketch could of course be of a generic pool and not refer in such specifics to the unit 4 pool.)

 

[joewein]

As an aside (see attachment) -- there is a square structure indicated in the drawing at what would be the east side of the unit 4 pool. I wondered if there is any significance to that depictured detail, as regards how the unit 4 pool set up is or is to be actually configurated.

I believe that's the skimmer tank. Every SFP has one.

To prevent any possibility of the pool draining from a ruptured pipe or from siphoning, there is no pipe connected to the bottom of the pool. Instead the water circulation system uses the skimmer tank such that when cold water gets added to the main pool it raises the level until it overflows into the skimmer tank, from where it is pumped into the heat exchanger (and from there back into the main pool).