Japan Earthquake: Nuclear Plants at Fukushima Daiichi

[MadderDoc]

From the descriptions in the accident analysis report, I get the impression that the SRVs talked about would be of a type which is called dual function direct acting valves. They are called dual function because they are able to be opened in two ways: a) automatically by the pressure reaching a set value (pressure mode), or b) by manual or remote actuation (actuation mode). They are called direct acting, because actuation operates directly on the valve body, not through an intermediate system.

I have attached a technical diagram of a dual function-direct acting valve cropped from
NRC report on SRV performance at US nuclear plants

In pressure mode, the valve would open automatically on exceeding the set point, which would in the case of Fukushima be around 7-8 MPa.

For manual or remote actuation, compressed air is needed to drive a pneumatic piston at the valve, and power is needed to open a valve that let compressed air flow to that piston. Then, the piston through a lever applies an opening force to the valve body, sufficient to open the valve -- on the condition that it can be assisted by an inlet steam pressure at or above about 0.7 MPa. If inlet pressure is below about 0.7 MPa. actuation will fail due to lack of opening force.

However, once successfully opened due to actuation -- (my question is, do I understand this correctly?) -- there would seem to be nothing that keeps the valve from closing except the pressure from the inlet steam. The piston-lever arrangement seems designed only to flip the valve open, not to keep it open. Once pressure falls below a certain level (about 0.3-0.4 MPa), the valve is meant to reseat, and to stay closed, and to open again only on actuation or by reaching its high pressure set point.

 

[SteveElbows]

A couple of possibly silly questions about detected level of hydrogen in reactor 2.

Firstly how come they don't seem to have increased the rate of nitrogen injection into the reactor beyond 5.0N m³/h for reactor 2? I am under the impression that they originally reduced it when they were getting things ready to do the endoscope investigation, and I think they turned the injection to PCV back up since then, but why they haven't done this with the reactor too?

Secondly the measured level of hydrogen seems to have ten increasing lately, and the 5am report for today seems to indicate its reached 0.42 vol %. What could explain this, and how high should the level reach before I express concern?

(Current data I am looking at is http://www.tepco.co.jp/nu/fukushima-np/f1/images/2012parameter/12042805_table_summary-j.pdf )

 

[zapperzero]

Does "cold shut down" also imply no active cooling required?

Nope.

 

[mheslep]

And? How far does the decay heat have to drop to eliminate the need for active cooling? Or more to the point, what is the Watts/deg C heat flux path to ambient?

 

[jim hardy]

@ madderdoc -

However, once successfully opened due to actuation -- (my question is, do I understand this correctly?) -- there would seem to be nothing that keeps the valve from closing except the pressure from the inlet steam. The piston-lever arrangement seems designed only to flip the valve open, not to keep it open.

did something written lead you to that thought?
It would be a surprise to me. It looks like a direct operated valve with a pneumatic cylinder to actuate the lever, which pulls directly on the spindle


If steam opens it it'll close when pressure comes down.
But if the handle opens it it'll stay open until handle is released. i think.


http://www.tycoflowcontrol.com/valves/Images/CROMC-0292-US.pdf

That bellows is not a pilot, it's there to make the valve insensitive to back pressure. see 2.4.3 of link

old jim

 

[MadderDoc]

@ madderdoc -
did something written lead you to that thought?

Yes I got that thought from reading the http://icanps.go.jp/eng/interim-report.html .
As attachment IV-6 it has a simplified diagram (see below) based on Tepco documents, explaining how SRVs work in actuation mode: (my boldface) "When nitrogen gas is fed into the cylinder, the piston and the stem are pushed up by the coupling lever. The valve body is then in a free state after the stem has been pushed up. When the valve body is pushed up by the steam pressure in this state, a steam flow channel is formed and steam is released into the S/C through the exhaust pipe."

In chapter IV of the report they explain further in footnotes, how much steam pressure at the inlet need to be in this 'free state' for a steam flow channel to actually result from an actuation (0.686 MPag) , and at which lower pressure the valve body from the open position will reseat by its own weight (0.345 MPag). (Edit: I would understand these figures to indicate the required differential pressure between inlet and outlet, rather than just the inlet pressure, but then I am just a chemist, not an engineer.)

 

[MadderDoc]

A couple of possibly silly questions about detected level of hydrogen in reactor 2.

Firstly how come they don't seem to have increased the rate of nitrogen injection into the reactor beyond 5.0N m³/h for reactor 2? I am under the impression that they originally reduced it when they were getting things ready to do the endoscope investigation, and I think they turned the injection to PCV back up since then, but why they haven't done this with the reactor too?

Perhaps they have other things they want done in Unit 2 or they want to make observation a while yet under the present degree of intervention -- in any case the job of the nitrogen injection is not to keep the hydrogen concentration low as possible, only to keep it at a safe level.

Edit: Just an afterthought, would one be able to say that "release of radioactive materials from the PCV is under control" while one would need to flush the PCV with huge amounts of N2 in order to keep other PCV parameters under control?

Secondly the measured level of hydrogen seems to have ten increasing lately, and the 5am report for today seems to indicate its reached 0.42 vol %. What could explain this, and how high should the level reach before I express concern?

In itself I can't see the current H2 level as a concern, it is more the fact that it has been rising significantly recently, and with no obvious explanation why. Explosive limit for hydrogen is about 4 %, but that is assuming there is oxygen, I don't know if the reactor atmosphere has any significant content of oxygen, I should hope not.

Current data shows a further increase over yesterdays value (0.44-->0.48 %) http://www.tepco.co.jp/nu/fukushima-np/f1/images/2012parameter/12042905_table_summary-j.pdf )

 

[jim hardy]

how much steam pressure need to be to be in this 'free state' for a steam flow channel to actually result from an actuation (0.686 MPag) , and at which lower pressure the valve body from the open position will reseat by its own weight (0.345 MPag).

Thanks !

is 0.686 mpa just 100 psi
and 0.345 mpa just 50 psi ?

sounds like just enough to lift the internal parts.


thanks again. Perusing part IV now.

 

[MadderDoc]

Thanks !

is 0.686 mpa just 100 psi
and 0.345 mpa just 50 psi ?

sounds like just enough to lift the internal parts.


thanks again. Perusing part IV now.

I hope you will return. I'd like to know, in Fukushima after the depressuring with such a valve -- on the assumption they kept the valve energized with power and compressed air throughout after that time, would that mean they upheld a pressure relief system that would make the RPV pressure cycle between 100, and 50 psi? And would that be psi gauge, or relative to the backpressure of the exhaust? (I imagine most parameters for such a system would be given for a situation where steam inlet pressure is very high and exhaust about at ambient, such as to moot the distinction, but that was not necessarily the situation in Fukushima)

Otoh, should in the midst of the explosions power or compressed air have failed them, wouldn't the valve have returned to normal mode, to open again only on re-actuation, or by the RPV pressures reaching high pressure set point, about 7-8 MPa, or 1000 psi?

 

[zapperzero]

Edit: Just an afterthought, would one be able to say that "release of radioactive materials from the PCV is under control" while one would need to flush the PCV with huge amounts of N2 in order to keep other PCV parameters under control?

Sure, why not. "Under control" does not mean "stopped".

In itself I can't see the current H2 level as a concern, it is more the fact that it has been rising significantly recently, and with no obvious explanation why. Explosive limit for hydrogen is about 4 %, but that is assuming there is oxygen, I don't know if the reactor atmosphere has any significant content of oxygen, I should hope not.

Of course there is oxygen. The limit is 4% in air, by the way. Inside the reactor there is not "normal" air, but rather air with a lot of added nitrogen (i.e. with a much smaller than normal proportion of oxygen).

Anyway, the H2 level should by all rights be rising, as they have decreased the rate of nitrogen injection.

 

[zapperzero]

And? How far does the decay heat have to drop to eliminate the need for active cooling?

IOW, used bundles are moved to dry cask storage after how long?

 

[MadderDoc]

<..>the H2 level should by all rights be rising, as they have decreased the rate of nitrogen injection.

But they have kept the reduced rate of nitrogen injection constant now for several weeks with H2 concentration steady at about 0.2 %. It is only during the last few days the H2 level has increased now to about 0.5 %.

 

[SteveElbows]

Firstly how come they don't seem to have increased the rate of nitrogen injection into the reactor beyond 5.0N m³/h for reactor 2? I am under the impression that they originally reduced it when they were getting things ready to do the endoscope investigation, and I think they turned the injection to PCV back up since then, but why they haven't done this with the reactor too?

Oops I just noticed that I got this the wrong way round, they had put the RPV nitrogen injection back to previous levels but its the PCV they left at only 5.0N m³/h

 

[r-j]

And? How far does the decay heat have to drop to eliminate the need for active cooling? Or more to the point, what is the Watts/deg C heat flux path to ambient?

In the US it is after 5 years in a cooling/storage pond. http://en.wikipedia.org/wiki/Dry_cask_storage#United_States

 

[jim hardy]

on the assumption they kept the valve energized with power and compressed air throughout after that time, would that mean they upheld a pressure relief system that would make the RPV pressure cycle between 100, and 50 psi?

indeed that's what is described by that attachment.
I am accustomed to valves where the stem lifts the disc directly, but this one may instead lift the spring which would allow the disc to behave as described.

And would that be psi gauge, or relative to the backpressure of the exhaust?

Yes, gage, that bellows keeps discharge pressure away from top of disc. So force up on disc is steam pressure X wetted area and force down is its weight plus pressure inside bellows (atmospheric - that's why bonnet is open type) plus spring force.

I'd like to find the Crosby manual for that specific valve. The link i gave is a bit generic.

should in the midst of the explosions power or compressed air have failed them, wouldn't the valve have returned to normal mode, to open again only on re-actuation, or by the RPV pressures reaching high pressure set point, about 7-8 MPa, or 1000 psi?

Yes, The piston would collapse back down to bottom of cylinder and allow spring to push stem down against disc.

That's how i see it.

old jim

 

[MadderDoc]

Decay heat question. It's estimated here that the decay heat from Daiichi-1 was down under 5MW after a couple months. Given an *undamaged* reactor and building, but still without power, what is the maximum heat level that can be rejected to ambient through passive means without uncovering the fuel? <..>

The reactor would need to be held at a relatively low temperature such as to avoid SRV's opening, and even lower than that, since we also don't want to degrade any part of the PCV. Just to pick a figure, let's say we think we can handle 127^oC or 400 K. A black-body of this temperature will radiate 5.67E-8*400⁴ watt/m². That's about 1.5 kW/m². The surface area of the RPV is in the neighbourhood of 500 m², so that would be 0.7 MW, as the maximum we would like to have in the RPV *if there were no containment around it*. But seeing there is, we would like to have less, and probably much less than 0.7 MW. I know this is not quite the figure you are asking for, but at least it puts a cap to it.

 

[jim hardy]

hmm two thoughts while i was eating breakfast.

1. That statement "..valve body is then in a free state.." is written by a technical writer for the report , i'd sure like to find it in the valve documentation. My skepticism is because i am accustomed to a direct plug-stem connection in the smaller safety valves in my experience. So i am allowing the possibility that's an error of translation by a bright fellow who's picked up a manual for pilot operated valve instead of that safety valve. If that's so, then lifting the valve handle would open the valve irrespectine of pressures. Hence my remark about finding the Crosby manual for that valve.
But for now i have to accept what's written.

2. I told you gage pressure. But - is not that valve located inside PCV ? So gage pressure in there is relative not to atmosphere but to PCV pressure.

The sketch you posted shows something called "Eductor" which i haven't figured out yet. I think its purpose is to reduce pressure above disc but outside bellows which helps give that snap-open characteristic you want in a safety valve.

I would like to more completely understand that valve's internals.
As an instrument guy i worked on regulating valves. Safety valves were in mechanical discipline and my small knowledge of them comes from talking with the mechanics who maintained them.
Ahh our regrets in life are mostly about the things we didnt learn. I could have learned more about code safeties.

old jim

 

[MadderDoc]

Thanks, that was a great help. Now I think I've got some thinking to do as regards the implications. Up till now I had the SRVs in Fukushima as something that was flipped to stay open, come what may. I can't remember reading about anything done with the SRVs after the time the reactors got depressurised.

 

[jim hardy]

Found some more Crosby literature.

The "Balancing Piston" is a backup in case the bellows ruptures.
A person accustomed to control valves might well interpret it as a pilot
and make the mistake i mentioned.

This link takes me to Crosby catalog 310 for their JOS and JBS series valves used on PWR's. Still trying to find exactly what's a 6R10.
http://www.google.com/url?sa=t&rct=j...eKGFteuMsdFGYg
page 59 describes the balancing piston.

And from a generic study of nuclear safety valves, page 11
http://www.osti.gov/bridge/purl.cover.jsp?purl=/41402-TpkPIC/webviewable/41402.pdf

The disc is the movable sealing element and seating
surface on which the following two forces act: (1) the
downward spring force transmitted by the spindle and
disc holder and (2) the upward force from the process
fluid pressure acting on the bottom surface of the disc.
The disc holder and spindle are held together by a
spindle retainer, and the disc is secured inside the disc
holder by a disc retainer, with both retainers generally
being snap rings.

and page 12:

Another component frequently found in PRVs is the
manual lift lever shown in Fig. 3.4. In nuclear
facilities this lever is wired down and not used by
operators in any procedures, but only by maintenance
personnel. The movement of the lever rotates the
lever shaft and lifting fork (or “dog”) that acts against
a release nut or load plate, thus lifting the spindle and
opening the valve.

page 4 has a photo of BWR safety. The piston that operates the handle is prominent.

Doc, i am ready to believe that "floating state" comment is an innocent error by technical writer.
That stem is i believe directly coupled to the disc, ie raising handle opens valve irrespective of pressure.

So by opening the valve they conneccted RPV to torus . I beieve. So long as their air and batteries held out.

old jim

EDIT sorry to flip on you
but one must go with best info available. I wasn't ready to accuse tech writer nased on my own limited experience. Apologies for the flip-flop.

 

[mheslep]

The reactor would need to be held at a relatively low temperature such as to avoid SRV's opening, and even lower than that, since we also don't want to degrade any part of the PCV. Just to pick a figure, let's say we think we can handle 127^oC or 400 K. A black-body of this temperature will radiate 5.67E-8*400⁴ watt/m². That's about 1.5 kW/m². The surface area of the RPV is in the neighbourhood of 500 m², so that would be 0.7 MW, as the maximum we would like to have in the RPV *if there were no containment around it*. But seeing there is, we would like to have less, and probably much less than 0.7 MW. I know this is not quite the figure you are asking for, but at least it puts a cap to it.

Thanks. This is the approach I'm looking for. Some questions:

Is the temperature of the PCV the critical figure for a walk away state, rather than the temperature of the fuel assembly? I assumed that on the low side no pressure was allowable in walk-away thus 100C was the limit internal to the PCV. Or, if a low steady-state pressure could be sustained indefinitely then something short of a temperature that trips the SRVs? I have no idea.

With regard to heat transfer, convective free air would dominate at about ~10W/m^2/K depending the humidity. If ~100K above ambient is allowable as you suggest, then ~1KW/m^2 is the convective heat transfer, so that a 500m^2 RPV allows .5 MW of decay power. For Daichi 2&3 I gather .5 MW is still many months away.

 

[mheslep]

IOW, used bundles are moved to dry cask storage after how long?

In the US it is after 5 years in a cooling/storage pond. http://en.wikipedia.org/wiki/Dry_cask_storage#United_States

The cask storage stage is the time required to allow deer and rabbits to nuzzle the outdoor casks without cooking them, or to prevent rain/snow from cracking an otherwise super heated cask. The PV (without active cooling) inside containment might sustain a somewhat more elevated power level might for a time without in a radiation release.

 

[MadderDoc]

Is the temperature of the PCV the critical figure for a walk away state, rather than the temperature of the fuel assembly? I assumed that on the low side no pressure was allowable in walk-away thus 100C was the limit internal to the PCV. Or, if a low steady-state pressure could be sustained indefinitely then something short of a temperature that trips the SRVs? I have no idea.

Thanks for the link.

Yes, I think the PCV temperature would be the critical figure. Something just short of a temperature that trips the SRVs would fry the PCV within hours, it only works under normal operation because the PCV is being cooled. With no power no cooling at hand it would be imperative to have the RPV and fuel at a temperature far below what they themselves can withstand. I'd agree that 100C could be the pain threshold for the PCV, rather than the figure I suggested, we also wouldn't like the PCV to loose steam.

 

[MadderDoc]

Doc, i am ready to believe that "floating state" comment is an innocent error by technical writer.
That stem is i believe directly coupled to the disc, ie raising handle opens valve irrespective of pressure.

So by opening the valve they conneccted RPV to torus . I beieve. So long as their air and batteries held out.

old jim

EDIT sorry to flip on you
but one must go with best info available. I wasn't ready to accuse tech writer nased on my own limited experience. Apologies for the flip-flop.

I really don't mind, jim, and I much appreciate your efforts to get it right, not just answered, also as long as I am getting wiser all the time, and I think I am.. :-)

Still most of the stuff about the technical workings of the valve is a bit over my capacity, but I do understand, that it is one crucial point whether that thing called the stem is directly connected to that thing called the disc. Assuming, as you, and everything else I have been able to dig up indicates, they are directly connected. So actuation of the valve would seem to imply that the valve would simply be forced open.

Otoh, the investigation committee who wrote it, and Tepco who I assume read it, have let pass in the interim report the considerations quoted below, which imply contrarily, that the SRV is not unconditionally forced open on actuation, so how to square the conflicting evidence? Could the situation be that the stem/disc action facilitates the opening of the valve, but that a certain steam pressure still would be needed to produce the flow channel?

"In general, the SRVs can be manually opened by remote control, if the RPV pressure is over 0.686MPa in gage. According to the plant parameters released by TEPCO, the Unit 3 RPV pressure at around 2:44 on March 13 was 0.580MPa in gage. Therefore the possibility that RPV pressure was below the required value at the time of the first opening operation at around 2:45 cannot be ruled out. On the contrary, taking into account a shift team operator’s logbook saying that the RPV pressure was 0.8 MPa at around 2:45 on the same day, it can be concluded that the lower pressure was not the real cause of the “fail to open.” To return to TEPCO’s plant parameters, the RPV pressure at Unit 3 around 3:00 on the same day elevated up to 0.770MPa in gage. If so, it is highly possible ...[snip]"

Edit: Jim, I just experienced a potential serendipity, please can I have you take a look page number 68 of the report and ff
http://pbadupws.nrc.gov/docs/ML1111/ML111170549.pdf
What is described there is technically way, way over my head, but just scanning the figures with my eyes
I get much too similar signals from this text as I get from the interim report to ignore.
Perhaps we are looking at Target Rock SRV's not Crosbys.

 

[jim hardy]

Thaks m'doc for the encoraging words.

that it is one crucial point whether that thing called the stem is directly connected to that thing called the disc. Assuming, as you, and everything else I have been able to dig up indicates, they are directly connected. So actuation of the valve would seem to imply that the valve would simply be forced open.

Otoh, the investigation committee who wrote it, and Tepco who I assume read it, have let pass in the interim report the considerations quoted below, which imply contrarily, that the SRV is not unconditionally forced open on actuation, so how to square the conflicting evidence?

I've found Dresser makes such valves and i found their catalog. It has nice color drawings that definitely show snap-rings . Will post link if i can find it again.

I have a personal dislike for pilot operated valves in this application. So i didnt look into Target Rock.

Crosby's drawings that I've found are lower resolution and don't seem to show a snap ring. In fact some of the drawings look like they'd agree with the TEPCO statement of floating disc.

After making that post i went outside to work on an old engine and thought about this.
I am biased by prior experience so tended to disbelieve the lack of firm stem-to-disc connection. That's my prejudice i realize now.
Given that safety valves evolved from a simple hole covered by a disc with weights stacked on top of it, there's no need for a solid connection to stem.
Indeed there's that historical pecedent.

So i answered before i should have. Realized that upon reflection.
However - the reason i gave answer i did is this:
Assume some reasonable diameter for the disc, say four inches. That gives it area of 4∏ square inches.
When pressure below disc X area of disc equals weight of parts to be lifted by steam , those parts will indeed be lifted by steam. (Assuming spring is held away by operating handle.)
Now 4∏ square inches X 100 psi is 1257 pounds. That's just too much for a disc and retainer to weigh. We're talking about something the size of harmonic damper on a big car engine.
At 2 inches it's still 314 pounds which pushes the credibility limit for me. I found this line in the 179549 link you gave:

The reactor vessel pressure muat
be at least 50 psi (0.345 MPa) above the wetwell pressure in order for
the main stage- to upen.

That quote is describing the Target-Rock as you suggested and it sure sounds a lot like what the TEPCO technical writer wrote.Target-rocks in my plant were maintenance headaches. And a pilot valve caused TMI. So i am biased against pilot valves for 'important to safety' service. I assumed the mechanical designers would stick to the simpler dierct acting design like Crosby.



another serendipity moment for you

Some operating BWRs are equipped with three-stage Target Rock valves, which have exhibited a greater tendency to stick open in the past than have other types of valves. Many BWR utilities, however, have replaced the original three-stage valves with the newer two-stage Target Rock valves (Figure 3.7-8). Some operating BWRs are equipped with Dresser electromatic relief valves. BWR-5 and BWR-6 plants are equipped with Crosby and Dikkers dual function SRVs (Figure 3.7-9).

Looks like the designers came around .

That's from NUREG/CR-6042 Rev. 2 section 3.7.2.5 , page 7 of 213
http://www.google.com/url?sa=t&rct=...rv6vifYChItD2Rh8g&sig2=ufvl8HNyM8y8zg5f6ovsUw

Fig 3.7-9 on page 33 is a better drawing of the Crosby type valve. Zoomiing into 300% i still can't tell how it's put together. It's frustrating - when i worked at plant we could go to warehouse and look at the on-hand spare parts.

But I'm coming to my senses now.
We don't know which valve they have.
So i withdraw my accusation against their tech writer, he may well have been describing a different valve than i was looking at. Apology to you, unknown writer.

And i back off my claim in last post. Go ahead with your thinking per TEPCO writer.

M'doc I admire you guys' doggedness and attention to detail. You're doing it right..


Imagine what Microsoft could have been if they had an industrial strength mindset.old jimPS - somebody has put pilot valves in main steam safety service. no comment.

 

[jim hardy]

From Crosby catalog 310 which claims to represent their Nuclear Main Steam Safety valves, probably JOS type:

The disc insert retention, disc holder and nozzle ring of the JOS-E
and JBS-E have been re-engineered to improve maintenance,
minimize spare parts and provide more component part
interchangeability. (Figure 2)
The disc insert is inserted into the disc holder with a retention clip
which is compressed as it passes through the smallest diameter in
the disc holder recess and then returns to its normal shape once it
has passed through. With the retention clip in its original shape,
the disc insert is held securely in place.

so it'll follow stem.

http://www.google.com/url?sa=t&rct=...7PhReZv1DNQdBw_zQ&sig2=r35JrFpoZNJuWyuJ3LE05Q

or CROMC-0297-US.pdf
text from page 4 and see the clip item 29 on page 7.

you're right on with that Target Rock. Wish i knew for sure what valves they have.

old jim

 

[MadderDoc]

<..> Wish i knew for sure what valves they have.

old jim

Yeah. Without that knowledge there may be left unresolved questions, but I think we are now better qualified to express them :-)

I'll assume that the SRVs they have, whatever their brand,
-- in pressure mode, are meant to open automatically at about 1000 psig (about 7-8 MPag), , and reclose at a pressure some 3-10% lower.
-- in actuation mode, are controlled by a differential pressure, which on exceeding 100 psi (0.686 MPa) will make the valve come open from its closed position,
and which on dropping below 50 psi (0.345 MPa) will make the valve come close from its open position.

It is not clear to me which differential pressure we are talking about. Some sources indicate steam inlet differential to PCV , some gauge, ie. steam inlet relative to ambient. Physically it seems the SRVs are inside the PCV, which under normal operation would be at ambient, so it would not matter much with that distinction, but it might under accident conditions. The more 'learned' the description, the more it seems to me indicated there that the PCV pressure factors into the differential pressure -- while the interim report in its considerations judges the available differential pressure only from the RPV gauge pressure. Only once, and only implicitly it says something to the effect that other factors might play a part (my boldface):

"The SRV, functionally, can be opened manually above the RPV pressure of 0.686MPa in gage and remain in an open position down to 0.344MPa in gage after the first actuation. Under this limiting pressure, however, the valve is to be fully closed because valve disk weight exceeds the lifting force. The SRV is, anyhow, less likely to be opened in the lower pressure ranges".

(The explicit specification 'first actuation' of the figures 0.686 MPa/0.344 Mpa would seem to imply that values for the second actuation are (edit: or could be) different)


Edit:
At the time of 'the first actuation' (Unit 3, in the morning of March 13th) the S/C pressure was 0.445 MPa. Assuming a Target Rock SRV with the properties described in the NUREG document this valve would come open when RPV is at 0.345 MPa above S/C pressure, i.e. at 0.790 MPa, or 0.690 MPag. This fits well with the figure 0.686 MPa gauge given by the interim report.

While being actuated, the Target Rock SRV main valve would be kept open by the differential pressure between the RPV and the S/C. IOW, when the RPV pressure would decrease to that of the S/C, 0.445 MPa, or 0.345 MPag, the main valve would come close. Again, this fits well with the figure 0.344 MPa gauge given by the interim report.

 

[jim hardy]

but it might under accident conditions. The more 'learned' the description, the more it seems to me indicated there that the PCV pressure factors into the differential pressure -- while the interim report in its considerations judges the available differential pressure only from the RPV gauge pressure.

Indeed the forces on the valve parts are difference between pressure inside and outside the valve. So if valve is inside PCV , outside of it sees PCV pressure. Or if a pilot valve it may see upstream(RPV) vs downstream(torus) pressures. Can I assume torus is about same as PCV pessure?

Now to RPV pressure:
If the pressure sensor is inside PCV and is a gage pressure sensor, same applies. It'll report difference between RPV and PCV, it knows nothing of atmosphere outside PCV and cannot compensate.
If the pressure sensor is an absolute pressure sensor that's what it will report, absolute pressure.

Gage and absolute pressure sensors are similar but absolute pressure ones are a bit more expensive. That's because they must include a sealed and evacuated chamber for an absolute zero pressure reference. Our Rosemount gage and absolute sensors were identical except that the gage ones leave that reference chamber open to local atmosphere.


Well thanks for the exercise ! Ilearned some things.

If your observed data is fitting with that Target-Rock valve model i'd say that clinches it - they have something similar.
If it really takes 50 psi differential to lift that plug then there's smaller unbalanced areas than i estimated from those Crosby drawings.

Nice work, Doc .
old jim

 

[NUCENG]

Indeed the forces on the valve parts are difference between pressure inside and outside the valve. So if valve is inside PCV , outside of it sees PCV pressure. Or if a pilot valve it may see upstream(RPV) vs downstream(torus) pressures. Can I assume torus is about same as PCV pessure?

Now to RPV pressure:
If the pressure sensor is inside PCV and is a gage pressure sensor, same applies. It'll report difference between RPV and PCV, it knows nothing of atmosphere outside PCV and cannot compensate.
If the pressure sensor is an absolute pressure sensor that's what it will report, absolute pressure.

Gage and absolute pressure sensors are similar but absolute pressure ones are a bit more expensive. That's because they must include a sealed and evacuated chamber for an absolute zero pressure reference. Our Rosemount gage and absolute sensors were identical except that the gage ones leave that reference chamber open to local atmosphere.


Well thanks for the exercise ! Ilearned some things.

If your observed data is fitting with that Target-Rock valve model i'd say that clinches it - they have something similar.
If it really takes 50 psi differential to lift that plug then there's smaller unbalanced areas than i estimated from those Crosby drawings.

Nice work, Doc .
old jim

The pressure difference between the PCV and the suppression chamber in a BWR is limited by vacuum breakers which relieve from the torus to the PCV if the torus pressure exceeds the PCV pressure. Other vacuum breakers relieve from the atmosphere to the torus air space to prevent either the torus or the drywell to become negatively pressurized compared to atmosphere (prevents the crushed beer can syndrome). The drywell pressure may exceed the torus pressure by the submergence head in the torus sowncomers. If the pressure exceeds that the water in the downcomers is displaced and the drywell relieves to the torus.

I know you are an instrument guy, so maybe I should not try to add to your question about pressure instruments. Forgive me if I misunderstood. In PWRs with large dry containments the pressure instumentation transmitters or sensors are usually inside the containment building but outside the shield wall. so they may be measuring gauge pressure to the containment. In BWRs the instrument transmitters or sensors are all located outside the PCV and can be identified as gauge or d/p by the fact that the d/p sensors have two lines and the gage instruments have one which means the difference is to secondary containment pressure. Your discussion of Rosemount pressure sensors is accurate about absolute pressure instrumentation.

 

[MadderDoc]

<..>Can I assume torus is about same as PCV pessure?

I think so, big differences between the drywell and the torus would not be expected, and in the measured data it also holds generally true. The one clear exception is Unit 2 during a period from shortly before midnight between March 14th and 15th, when drywell pressure far exceeded torus pressure.

<..>
If your observed data is fitting with that Target-Rock valve model i'd say that clinches it - they have something similar.

I wouldn't say clinched, but something like that model is the only explanation that does not seem to lead to inconsistency. And from what you dug up it would be very much what one would've expected to find in this type of reactor.

If it really takes 50 psi differential to lift that plug then there's smaller unbalanced areas than i estimated <..>

I think there could be a main valve preload spring to determine the differential (see attached, from the GE Manual). So, when the interim report says it is reseated by its own weight, that does _not_ fit the Target Rock SRV. However, knowing about the preload spring is not helpful to the understanding of the basic principle of the valve, plausibly the originator of the text has left it out as just a distracting detail.

 

[MadderDoc]

<..> In PWRs with large dry containments the pressure instumentation transmitters or sensors are usually inside the containment building but outside the shield wall. so they may be measuring gauge pressure to the containment. In BWRs the instrument transmitters or sensors are all located outside the PCV and can be identified as gauge or d/p by the fact that the d/p sensors have two lines and the gage instruments have one which means the difference is to secondary containment pressure. Your discussion of Rosemount pressure sensors is accurate about absolute pressure instrumentation.

I've not caught the interim in being vague or inconsistent in their pressure expressions, when gauge is meant it is consistently specified, and consistently applied, and defined as meaning absolute pressure minus 0.101 MPa.

I think in the context of their investigation and report, for ease the investigation committee looked at the PCV pressures at the time relevant for their considerations of SRV operations, and they found the S/C pressure at those periods of interest to have been generally about 0.44 MPa.

They could then determine -- at conditions of SC pressure 0.44 MPa -- what the reactor pressure would need to be in order for actuation to result in the opening of the valve. They found that figure to be about 0.69 MPag, (RPV pressure is usually measured and reported as gauge pressure, while PCV pressure is read and reported as absolute pressure).

Knowing that 0.69 MPag was the approximate limit, it could then be judged directly from measured data of the RPV pressure, whether or not at the time of a particular measurement an actuation of the SRV would have been expected to succeed, or to fail due to the RPV's not being at a sufficient pressure above that of the S/C.

 

[jim hardy]

Thanks Nuceng

In PWRs with large dry containments the pressure instumentation transmitters or sensors are usually inside the containment building but outside the shield wall. so they may be measuring gauge pressure to the containment.

Exactly how mine was built (Westinghouse 3 loop). Sensors were gage.
I do not recall whether the subcooled margin monitor used gage or absolute sensor. I do recall raising that question in design review.

[QUOTEIn BWRs the instrument transmitters or sensors are all located outside the PCV and can be identified as gauge or d/p by the fact that the d/p sensors have two lines and the gage instruments have one which means the difference is to secondary containment pressure. ][/QUOTE]

I did not know that. I assumed they'd be inside drywell.
Sum total of my BWR experience is a very brief (like an hour) tour of Duane Arnold plant thirty+ years ago.

If i learn something every day , and can turn things around so (rate of absorb) > (rate of forget), i may know something someday.

Thanks !



old jim

 

[jim hardy]

I think there could be a main valve preload spring to determine the differential (see attached, from the GE Manual).

Great drawing.

Bottom of piston can be assumed at upstream pressure
and top of piston at either upstream or downstream pressure depending whether pilot is in left or right position.

So:
Free Body Diagram when pilot is in right (open) position:
Force up = (area of piston bottom) X Pupstream + (area of seat) X Pdownstream
Force down = (area of piston top) X Pdownstream + (area of seat) X Pupstream + spring preload

when those two are equal valve can open.

From your drawing (presumably unscaled) - on my screen;
diameter of seat = 0.569 inch
diameter of piston = 1.025
diameter of shaft = 0.350

Let's see if we can "normalize" as nuclear egineers love to do:

diameters relative to diameter of seat:
diameter of seat = 0.569 inch / 0.569 = 1
diameter of piston = 1.025 / 0.569 = 1.801
diameter of shaft = 0.350 / 0.569 = 0.615

and areas relative to area of seat
area of seat =( 0.569 inch / 0.569)^2 = 1,
area of piston = (1.025 / 0.569)^2 = 1.801^2 = 3.243
area of shaft = (0.350 / 0.569)^2 = 0.615^2 = 0.378
area of piston bottom = 3.243 - 0.378 = 2.865

oksy it's all ratio'ed to seat area so we won't need ∏r^2 for a while

now back to free body diagram

Force up = Pupstream X (area of piston bottom) + Pdownstream X area of seat
Force down = Pdownstream X (area of piston top) + Pupstream X (area of seat) + spring preload

By assuming Pdownstream = zero(gage) we can make it a lot simpler

Force up = Pupstream X (area of piston bottom )
and
Force down = Pupstream X (area of seat) + spring preload

Equating those two
Pupstream X (area of piston bottom ) = Pupstream X (area of seat) + spring preload
Pupstream X (area of piston bottom - area of seat ) = spring preload

Valve can open when
Pupstream = spring preload / (area of piston bottom - area of seat )
Pupstream = spring preload / ((2.865 - 1)(area of seat))

Pupstream = spring preload /(1.865 X area of seat)

If it takes 50 psi to open valve
50 X 1.865 = springload/ area of seat
93.25 X area of seat = spring preload

So a seat area of say 4∏ sq inches (wild-a** guess at 4 inch disc) requires but a 1171 pound spring preload.
That still sounds high to me
but compare that spring to the one required for a direct acting valve like the Crosby

4∏ X 1100 psi = 13,283 pounds and that's a really stout spring. Pilot design allows 10X reduction in spring .

Oops i neglected weight of piston in above. Surely it's less than spring preload in case valve is installed upside-down.

Sorry for the digression - it do this for my own sanity checks. If it has entertained you it was worthwhile.
Corrections or suggestions welcome. I feel like there's an arithmetic mistake in it
but got to go now.
Would you guess that spring might be about size of an automobile suspension coil spring?

later
old jim

 

[MadderDoc]

<..> I feel like there's an arithmetic mistake in it
but got to go now.
Would you guess that spring might be about size of an automobile suspension coil spring?

I thought they were smaller , and it is by far not the first time I have underestimated the size of things in a nuclear plant! That was enjoyable, also to see how you tackled the problem with that unifying approach. I could find nothing wrong with the arithmetic. :-)

 

[MadderDoc]

Unit 2 March 14th - March 15th

The behaviour of pressures of the RPV and PCV in Unit 2 after it was depressurised at about 18:00 on March 14th
is grossly inconsistent with the hypothesis that the RPV was henceforth held depressurised by way of the relief valve.
The hypothesis should be rejected.

This hypothesis would predict that the RPV pressure was held at a level at about 0.35 MPa above that of the wetwell,
and the RPV pressure was by far not held at such level.

After the initial depressurizing followed a short period with the expected behaviour, but then RPV pressure was seen rising and falling
in three large peaks that went way outside what an actuated SRV would allow.

Concurrent with the second and largest of these RPV pressure excursions,
the drywell pressure (as well as the drywell CAMS reading, not shown in the figure) rose dramatically.

After the third and last RPV pressure excursion, the RPV and the drywell pressure equilibrated at level more than 0.5 MPa above that of the wetwell. This state was upheld until the morning of March 15th, when, concurrent with the onset of steaming from the reactor building, the pressures of the RPV and the drywell decreased, while the wetwell pressure went downscale.

Seeing the hypothesis that there was pressure relief through a SRV is not viable, and the RPV pressure did nonetheless not increase to hit the roof (automatic safety valve opens at 7 Mpa), the RPV must have found or made itself other channels for pressure relief during this period. IOW, most plausibly, the RPV was damaged.

 

[MadderDoc]

Unit 3 March 13th - March 15th

In unit 3, during the period after it was depressurised in the morning of March 13th,
a prominent feature of the RPV and PCV pressures
is that their variations henceforth appear to have been very closely correlated.
However while drywell and wetwell also appear to have been close to pressure equilibrated,
the RPV pressure tracked their variations only from a lower pressure level.

The RPV's being at a lower pressure than the PCV is difficult to explain, except by assuming that the RPV pressure readings are in error, in the sense: affected by a systematic error producing too low readings. With the added assumption that the barrier between the atmospheres of the RPV and the PCV had degraded, it could be explained how all three compartments would then effectively have been pressure equilibrated.

A hypothesis that the SRV was held open over the period fails, for a surprising reason: there is no indication that the RPV after its initial depressurising ever again attained a pressure of 0.345 MPa above the pressure of the wetwell, and therefore the main valve of the SRV would have come shut shortly after the initial depressurizing, and would have remained closed, whether or not the actuation of the valve was maintained.

Relief from the system had then to be by other means. There are in fact known S/C vents at about 42 and 46-47 hours after the earthquake, concurrent to peaks in the graph. Further opening and closing of S/C vents have been assumed for the models such as to fit the variations in pressure that followed -- except for the abrupt relief at 68 hours, the occurrence of which is so far unexplained by the models.

 

[jim hardy]

Did those show up in jstolfi's (remarkable) plots ?

This one shows rpv tracking drywell with the offset you noted:
http://www.ic.unicamp.br/~stolfi/EXPORT/projects/fukushima/plots/cur/out/pcor-PCA-PD-un3-full.png

i don't know what he used for time zero.


Maybe nuceng knows where that "A" reactor pressure sensor's tap is located.
I have been suspect of its readings since 3/21/11 when it went impossibly high then cured itself.
That shows on another of jstolfi's wonderful plots:
http://www.ic.unicamp.br/~stolfi/EXPORT/projects/fukushima/plots/cur/out/plot-pres-un3-t-I-full.png
it's too wide for page or i'd post it

and
http://www.nytimes.com/2011/03/21/world/asia/21japan.html?pagewanted=all

The Tokyo Electric Power Company, which runs the plant, appeared to have experienced a serious setback as officials said that pressure buildup at the ravaged No. 3 reactor would require the venting of more radioactive gases.

But at a news conference a few hours later, officials from the power company said that the pressure had stabilized and that they had decided they did not need to release the gases immediately, which would have heightened worries about wider contamination among the population. They said they were unsure what had caused the pressure to rise, highlighting the uncertainty engineers must still grapple with at Fukushima.

just one of those little nagging questions. This obsessiveness is part of my aspergers i think !

old jim

 

[MadderDoc]

Did those show up in jstolfi's (remarkable) plots ?

This one shows rpv tracking drywell with the offset you noted:
http://www.ic.unicamp.br/~stolfi/EXPORT/projects/fukushima/plots/cur/out/pcor-PCA-PD-un3-full.png
i don't know what he used for time zero.

Jorge uses midnight between March 10 and March 11 as time zero, and includes a much longer time series than the one I have been focusing on. As regards the abrupt relief at 68 hours (Jorge 82 hours) that has been left unexplained by the model, it may be a hint that 68 hours is also the time of the explosion of Unit 3.

Maybe nuceng knows where that "A" reactor pressure sensor's tap is located.
I have been suspect of its readings since 3/21/11 when it went impossibly high then cured itself.
That shows on another of jstolfi's wonderful plots:
http://www.ic.unicamp.br/~stolfi/EXPORT/projects/fukushima/plots/cur/out/plot-pres-un3-t-I-full.png
it's too wide for page or i'd post it

and
http://www.nytimes.com/2011/03/21/world/asia/21japan.html?pagewanted=all

I think Tepco would have been referring to the gradual build up of pressure that was seen at that time. That build up of pressure levelled out at not too critical levels, and Tepco canceled intervention, it seems, and all the while pressure started to drop back again. It was on the decreasing flank of that pressure variation that one of the RPV sensors returned very high readings (~10 MPa) for an hour or so, cause unknown. However, that was well after Tepco's consideration to vent, and also after the cancellation announcement, so probably unrelated to those matters. There could be reason to think something untoward happened (on top of all the untoward that had already happened), in connection with that apparent transient -- in data it appears as being the last throes of the reactor as a pressure containing system.

just one of those little nagging questions. This obsessiveness is part of my aspergers i think !

old jim

Lol. Just keep in mind, paraphrasing the proverb: "Those who restrain obsession, do so because theirs is weak enough to be restrained" :-)

 

[jim hardy]

The RPV's being at a lower pressure than the PCV is difficult to explain, except by assuming that the RPV pressure readings are in error, in the sense: affected by a systematic error producing too low readings.

in measuring low pressures the relative elevations of measurement points becomes significant. The offset you noticed , ~90kpa(?) equates to ~30 feet of water.

But it takes someone who knows the physical plumbing runs to know where the measuring taps are located. And whether there's a long vertical run of pipe in the line between sensor and mesauring point. Usually pressure sensors are calibrated to include head due to full sense lines and if that fill is lost (eg boils away when containment is hot and pressure is low) , the reported reading will be in error by amount of fluid lost.

Again these are just little details to be uncovered . Your observations are right on.
I did a lot of troubleshooting in my day . As the increasingly fine details emerge these little questions all resolve and a true picture emerges. But it sure morphs a lot along the way.

Nice work !

It was on the decreasing flank of that pressure variation that one of the RPV sensors returned very high readings (~10 MPa) for an hour or so, cause unknown.

~10mpa is ~1400 psi. As i say I'm just waiting for that detail to unravel . I believe it was a measurement error and I've not heard it explained.
But you and Elbows are much better versed than i am. If you run across a "why" for that one, please post.

old jim

PS - Thanks for indulging my OCD.

 

[MadderDoc]

in measuring low pressures the relative elevations of measurement points becomes significant. The offset you noticed , ~90kpa(?) equates to ~30 feet of water.

But it takes someone who knows the physical plumbing runs to know where the measuring taps are located. And whether there's a long vertical run of pipe in the line between sensor and mesauring point. Usually pressure sensors are calibrated to include head due to full sense lines and if that fill is lost (eg boils away when containment is hot and pressure is low) , the reported reading will be in error by amount of fluid lost.

The instrument would then read out too high values, if I understand the setup. Then of course, it could be the measured PCV pressure readings in unit 3 which are erroneously too high during this period, rather than the RPV's pressure readings too low.? At least I cannot now exclude that possibility, that may come later. As you say, it would not be unwelcome at all having some input from a BWR guy.

 

[jim hardy]

The instrument would then read out too high values, if I understand the setup. Then of course, it could be the measured PCV pressure readings in unit 3 which are erroneously too high during this period, rather than the RPV's pressure readings too low.?

well - if the pressure tap is above the transmitter
the transmitter would be calibrated to report less pressure than it sees.
That's because the condensed water in the vertical line adds to the pressure as you traverse down it. Kirchoff's pressure law ?


So if the pressure tap were near top of vessel and the sensor lower than that,
when sensing line dried out,
reported pressure would be low by the height of fluid lost.

but i don't know physical arrangement in a BWR. Mine i knew pretty well.

old jim

 

[MadderDoc]

What, if anything, could have been a _technical_ rationale of Tepco of letting firefighters risk their lives to splash water to the top of Unit 3 in the days following March 16th 2011?

 

[zapperzero]

What, if anything, could have been a _technical_ rationale of Tepco of letting firefighters risk their lives to splash water to the top of Unit 3 in the days following March 16th 2011?

The real and present danger of pool dry-out of course. We've been over this before, if from a different angle - the firefighting helicopters found multi-Sievert/h fields above unit 3, suggesting either massive gas/particulate releases (unlikely, for a variety of reasons) or shine from a mostly empty pool.

Of course, I don't see TEPCO ready to admit that yes, they almost lost Tokyo.

 

[zapperzero]

Then of course, it could be the measured PCV pressure readings in unit 3 which are erroneously too high during this period, rather than the RPV's pressure readings too low.

Common sense would suggest that whichever is closest to atmospheric is true, fwiw.

 

[MadderDoc]

The real and present danger of pool dry-out of course. We've been over this before, if from a different angle - the firefighting helicopters found multi-Sievert/h fields above unit 3, suggesting either massive gas/particulate releases (unlikely, for a variety of reasons) or shine from a mostly empty pool.

Those high radiation fields were measured, I believe, on March 16th, on which day there was massive steam release from the building -- Tepco evacuated their own people temporarily out of fear that the PCV was leaking, and videos were taken from helicopters showing steam sources to be consistent with that fear. Otoh, there does not seem to have been any cause on that day to think the pool was even boiling.

Therefore I do not find it readily credible that Tepco the next day should have thought the danger of dry-out of the pool of unit 3 was real and present to a degree of imminence that the brave people of the civil forces should think they had cause to be willing to risk their lives in desperate measures to splash just some water into it.

 

[MadderDoc]

Common sense would suggest that whichever is closest to atmospheric is true, fwiw.

Please explain.

 

[zapperzero]

Those high radiation fields were measured, I believe, on March 16th, on which day there was massive steam release from the building -- Tepco evacuated their own people temporarily out of fear that the PCV was leaking, and videos were taken from helicopters showing steam sources to be consistent with that fear. Otoh, there does not seem to have been any cause on that day to think the pool was even boiling.

Therefore I do not find it readily credible that Tepco the next day should have thought the danger of dry-out of the pool of unit 3 was real and present to a degree of imminence that the brave people of the civil forces should think they had cause to be willing to risk their lives in desperate measures to splash just some water into it.

They weren't expecting that water to make it into the PCV, no? So what else is left? A fire (for which there is no evidence whatsoever) or a pool that is _believed_ to be emptying fast.

 

[MadderDoc]

They weren't expecting that water to make it into the PCV, no? So what else is left? A fire (for which there is no evidence whatsoever) or a pool that is _believed_ to be emptying fast.

But you can't just assume that they did have a technical rationale relating to conditions of the plant. It is well known that people may act out of a variety of motivations. For what we know there was as little evidence that the pool was emptying fast, as there was of a fire.

 

[zapperzero]

Please explain.

I'd expect that one vent path or another eventually opened and remained open, so I'd intuitively go for the lower number, when asked to estimate which of the three readings of our steaming teapot is correct.

Moreover, the three sensor readings track each other pretty nicely - except at the initial two peaks, where the RPV pressure is shown as highest (which makes sense, as it could probably still hold pressure at those times). What seems more likely? That the RPV pressure is over-, then under-reported by the same instrument, or that the other two are consistently over-reported?

 

[zapperzero]

But you can't just assume that they did have a technical rationale relating to conditions of the plant.

Why not? Everything else, including the aborted decision to abandon the site, was based on a technical rationale.

It is well known that people may act out of a variety of motivations. [/ quote]

I don't really buy the "they did this to be seen doing something" idea anymore. Something WAS being done, quite arduously and visibly in fact. Trucks and pumps were moving in, power lines were being repaired etc etc.

For what we know there was as little evidence that the pool was emptying fast, as there was of a fire.

No. There was a simple calculation as to how long it takes for the pool to boil after the pump stops. There was also an elevated and increasing, but rather uniform radiation field around Unit 3 (on the ground), which at least to some must have said"gamma shine".

 

[MadderDoc]

Why not? Everything else, including the aborted decision to abandon the site, was based on a technical rationale.

:-) That is some statement but even if assumed it doesn't follow that this one would have to be.

I don't really buy the "they did this to be seen doing something" idea anymore. Something WAS being done, quite arduously and visibly in fact. Trucks and pumps were moving in, power lines were being repaired etc etc.

Now I didn't suggest Tepco did anything to be seen doing something. Perhaps they did, as I said people are known to act from a diversity of motivations, including this one.

There was a simple calculation as to how long it takes for the pool to boil after the pump stops.

Yes, and no doubt Tepco did the calculation, only the result must've shown that the pool ought to be days from boiling and many days from fuel exposure. You also never heard a Tepco official say the Unit 3 pool was drying out fast. Nor was it expressed at the next level of the foodchain, NISA, who rather thought at the time that the Unit 4 pool was the one in need of urgent attention.

If Tepco really thought the situation of the unit 3 pool was so desperate, then why did they clear the ground at unit 4 for a concrete pump asap, while it took them days to not even make a passageway at unit 3 such that the rescue teams could approach the pool?

There was also an elevated and increasing, but rather uniform radiation field around Unit 3 (on the ground), which at least to some must have said"gamma shine".

To be sure the dose rate was high around unit 3 after the explosion. That just puts some seriousness to the question, what was Tepco's technical rationale for letting people go in there to splash water onto the building.