Help | Contact Us
NukeWorker.com
NukeWorker Menu TMI 2 Question

Author Topic: TMI 2 Question  (Read 11453 times)

0 Members and 1 Guest are viewing this topic.

Offline nspunx4

TMI 2 Question
« on: Mar 02, 2012, 11:47 »
Hello,

First I would like to thank everyone for all of the knowledge I have received from this forum. I am just an interested "civilian" who enjoys learning about nuclear power.

My question is this... Can anyone explain why all of the official reports on the TMI 2 incident state that the emergency feed water block valves being closed for the first 8 minutes into the transient had no direct effect on the outcome other than to possibly distract the operators? From my very rudimentary and uneducated understanding of a PWR I thought the introduction of cold emergency feed water at the start of the loss of main feed water would have caused the steam generators to accept more heat from the RCS and kept the pressure from increasing thereby causing the pilot operated relief valve on the pressurizer to remain closed?

I am sure I am missing something here possibly the effect of the once through steam generator design B&W used?

Thank you in advance!

Offline jams723

Re: TMI 2 Question
« Reply #1 on: Mar 03, 2012, 12:17 »
First, due to the transient, the PORV open on the initial overpressure and failed to fully reseat.  Because the design had other high pressure pumps that were injecting cool water into the reactor vessel the EFW would have been beneficial but not having it at first did not have any effect on the transient.  The B&W design at that time was to allow the PORV valve to cycle.  You also have to avoid having too much heat transfer which is overcooling and has it's own issues.  EFW is very effective on the once through steam generator.

Bottom line is they had a pressurizer steam space Leak (the PORV), did not recognize it and turned off the high pressure injection pumps... That is the point when the EFW Block valves being closed affected the transient.

Offline nspunx4

Re: TMI 2 Question
« Reply #2 on: Mar 03, 2012, 12:35 »
I think I understand now. The initial over pressure condition would have been present with or without immediate introduction of EFW. Thank You!

thenuttyneutron

  • Guest
Re: TMI 2 Question
« Reply #3 on: Mar 03, 2012, 10:05 »
Assuming the PORV reclosed after it lifted the first time and they had EFW, this disaster would not have occurred.  B&W plants are very sensitive to a loss of feedwater event.  There are only about 40 gallons per inch in the steam generator and you need several hundreds of gallons per minute to remove the decay heat of the reactor (~250 gpm 10 min post trip).

The loss of subcooling margin is very serious for a PWR.  As long as the Reactor Coolant system is subcooled, adequate core cooling is assured.  The PORV being stuck open did not cause the core melt.  The fact that they overrode safety systems that had automatically started as designed by stopping them is where they went bad.  Even after all of this, the core was probably ok.  The final hit was when they tripped the Reactor Coolant Pumps and removed all heat removal from the core.

I think I understand now. The initial over pressure condition would have been present with or without immediate introduction of EFW. Thank You!

I am not sure if this is correct.  I am betting that the trip and initial SG blow down did not lift the PORV.  The PZR Spray should have been enough to stay away from the PORV lift setpoint.  This is all just an educated guess from me and should in no way be construed as a fact.
« Last Edit: Mar 03, 2012, 10:09 by Nutty Neutron »

Fermi2

  • Guest
Re: TMI 2 Question
« Reply #4 on: Mar 03, 2012, 11:44 »
Actually in the 70s a turbine/reactor trip on a BW tended to operate the PORVS 100% of the time.

Offline nspunx4

Re: TMI 2 Question
« Reply #5 on: Mar 03, 2012, 11:58 »
Thank you for the great responses. What changed since the 1970's to cause less cycling of the PORV's? Did the operators having the pressurizer spray on manual control have any effect? How long would the main feed pumps been able to circulate a 2 phase mixture before suffering catastrophic failure?

I find this technology to be really fascinating and want to understand as much as possible. There's only so much you can learn from the books and web sites without ever actually "being there" so to speak so any insight that any of you would be willing to share regarding this topic is greatly appreciated.

thenuttyneutron

  • Guest
Re: TMI 2 Question
« Reply #6 on: Mar 03, 2012, 12:14 »
Actually in the 70s a turbine/reactor trip on a BW tended to operate the PORVS 100% of the time.

I could see this being the case because in the old days, the plant was supposed to handle a turbine trip at 100% power without leaving Mode 1.  I can ask a fellow RO on my shift that has had his license longer than I have been alive.

Thank you for the great responses. What changed since the 1970's to cause less cycling of the PORV's? Did the operators having the pressurizer spray on manual control have any effect? How long would the main feed pumps been able to circulate a 2 phase mixture before suffering catastrophic failure?

I find this technology to be really fascinating and want to understand as much as possible. There's only so much you can learn from the books and web sites without ever actually "being there" so to speak so any insight that any of you would be willing to share regarding this topic is greatly appreciated.

The spray valve switch that I am used to seeing has three positions; closed, auto, and open.  The Spray valve switch is in auto with the valve in the closed position unless something goes wrong (e.g. RCS pressure input into the spray valve logic fails high).  The spray valve that I have only goes to 40% open in auto but full 100% if the switch is placed in the open position.  Closing the Spray Valve is one of my immediate actions to perform if the Spray Valve opens on me when it is supposed to be in auto & closed.  I have never placed the spray valve in the "closed" position at power.  I have however closed it while in Mode 3 to setup the spray bypass flow prior to going to Mode 2.

If the ROs had manually sprayed the PZR at the start, I think that it is possible that they would not have lifted the PORV.  During a plant trip caused by a loss of feed water, worrying about the spray valve should not be a concern at that point.  You have to get through your immediate actions of verifying that the RX is tripped before going to the other stuff.

TMI is a PWR and the Main Feed Pumps do not "recirculate" a two phase fluid for core cooling.  They pump subcooled water to the shell side of the steam generators.  If they are pumping a two phase fluid, that is called cavitation.  They won't last long and most likely will be tripped automatically or by the operators.
« Last Edit: Mar 03, 2012, 12:33 by Nutty Neutron »

Offline Starkist

Re: TMI 2 Question
« Reply #7 on: Mar 03, 2012, 12:24 »
How long would the main feed pumps been able to circulate a 2 phase mixture before suffering catastrophic failure?


This answer is "it depends". Factors include design of the pump, temperatures/pressures  ergo -> quality factor of the mixture, etc.

Some pumps are designed to cavitate without breaking (almost to an extreme), some shatter to pieces at the first sign of cavitation. They are usually over built pumps, so they'd most likely be able to last for any transient.   

Offline nspunx4

Re: TMI 2 Question
« Reply #8 on: Mar 03, 2012, 12:49 »
Im sorry I meant to say RCS not MFP above about pumping a water and steam mixture. I would imagine that if you had a low pressure condition in the RCS causing steam bubbles to form in the coolant then you would have to resort to a feed and bleed type of operation with injection pumps and let down to maintain some sort of core cooling or would you use the injection pumps to repressurize the RCS and collapse the steam bubbles?

Offline jams723

Re: TMI 2 Question
« Reply #9 on: Mar 03, 2012, 12:55 »
Actually in the 70s a turbine/reactor trip on a BW tended to operate the PORVS 100% of the time.

That is correct, the set point for the PORV was lower as it was designed to be used in transient response... Not that way now.

thenuttyneutron

  • Guest
Re: TMI 2 Question
« Reply #10 on: Mar 03, 2012, 01:01 »
Im sorry I meant to say RCS not MFP above about pumping a water and steam mixture. I would imagine that if you had a low pressure condition in the RCS causing steam bubbles to form in the coolant then you would have to resort to a feed and bleed type of operation with injection pumps and let down to maintain some sort of core cooling or would you use the injection pumps to repressurize the RCS and collapse the steam bubbles?

My job 1/2 the time is to be a spectator ready trip the Reactor Coolant Pumps on a loss of Subcooling Margin (SCM).  This is performed to limit the amount of RCS inventory that I lose during a LOCA with a loss of SCM.

You maintain full injection flow until SCM is reestablished.  Once you have SCM, you throttle the injection to keep SCM at a minimum.  You don't want a brittle fracture from pumping in cold water and over pressurizing the RCS.  

For a feed and bleed operation, you lock open the PORV and inject water in.  To collapse the steam bubbles, get EFW going for a boiler/ condenser type cooling.
« Last Edit: Mar 03, 2012, 01:04 by Nutty Neutron »

Offline jams723

Re: TMI 2 Question
« Reply #11 on: Mar 03, 2012, 01:07 »
Assuming the PORV reclosed after it lifted the first time and they had EFW, this disaster would not have occurred.  B&W plants are very sensitive to a loss of feedwater event.  There are only about 40 gallons per inch in the steam generator and you need several hundreds of gallons per minute to remove the decay heat of the reactor (~250 gpm 10 min post trip).

The loss of subcooling margin is very serious for a PWR.  As long as the Reactor Coolant system is subcooled, adequate core cooling is assured.  The PORV being stuck open did not cause the core melt.  The fact that they overrode safety systems that had automatically started as designed by stopping them is where they went bad.  Even after all of this, the core was probably ok.  The final hit was when they tripped the Reactor Coolant Pumps and removed all heat removal from the core.

I am not sure if this is correct.  I am betting that the trip and initial SG blow down did not lift the PORV.  The PZR Spray should have been enough to stay away from the PORV lift setpoint.  This is all just an educated guess from me and should in no way be construed as a fact.

Not quite accurate, without the RCPs running they still could have avoided core melt by running the makeup pumps which inject essentially in the core... This is known as HPI cool down, or using EFw set up natural circulation and performed a cool down that way. It was stopping the HPI pumps which stopped putting cool water into the core and stopped keeping the core covered.

Additionally, the B&W once through steam generator has very little inventory to blow down (unlike a u-tube S/G).  With the way TMI tripped as a loss of feed resulting in rx heat generation continuing for a short period of time, the high pressure would have exceeded the spray valve capacity and reached the PORV set point.

thenuttyneutron

  • Guest
Re: TMI 2 Question
« Reply #12 on: Mar 03, 2012, 01:11 »
Not quite accurate, without the RCPs running they still could have avoided core melt by running the makeup pumps which inject essentially in the core... This is known as HPI cool down, or using EFw set up natural circulation and performed a cool down that way. It was stopping the HPI pumps which stopped putting cool water into the core and stopped keeping the core covered.


Everything that I have been taught says that the 2 phase flow across the core kept it from being damaged.  When they stopped the RCPs due to cavitation, the flow stopped and the phases separated thus allowing the fuel to be damaged.  Either way, the pumps probably would have gone away on their own had they kept losing the RCS inventory and the end result would be the same.

I think the way we are arguing here is about like arguing that the suicidal jumper died from the jump vs the sudden acceleration at the end.  Either way the result is the same.

They screwed up bad when they blocked and stopped the injection systems, but the core had not been damaged yet.
« Last Edit: Mar 03, 2012, 01:21 by Nutty Neutron »

Offline jams723

Re: TMI 2 Question
« Reply #13 on: Mar 03, 2012, 01:17 »
I disagree.  Everything that I have been taught says that the 2 phase flow across the core kept it from being damaged.  When they stopped the RCPs, the flow stopped and the phases separated thus allowing the fuel to be damaged. 

Then we can agree to disagree. The reason they got to 2 phase flow was because they turned off the HPI pumps due to being trained to never go solid.  If they had not done that, there never would have been core damage.  I agree in that because they did turn off the HPI pumps and placed themselves in the situation where they turned off RCPs that ultimately resulted in core damage.  It is debatable whether the RCPs would have continued to run even if they had left them in operation.

In my opinion it is like a root cause analysis... Did securing the RCPs result in a contributing cause... Yes.... But failure to keep injecting water.... Not keeping the core covered was the root cause.

Fermi2

  • Guest
Re: TMI 2 Question
« Reply #14 on: Mar 04, 2012, 09:44 »
Then we can agree to disagree. The reason they got to 2 phase flow was because they turned off the HPI pumps due to being trained to never go solid.  If they had not done that, there never would have been core damage.  I agree in that because they did turn off the HPI pumps and placed themselves in the situation where they turned off RCPs that ultimately resulted in core damage.  It is debatable whether the RCPs would have continued to run even if they had left them in operation.

In my opinion it is like a root cause analysis... Did securing the RCPs result in a contributing cause... Yes.... But failure to keep injecting water.... Not keeping the core covered was the root cause.

Agreed.

Offline nspunx4

Re: TMI 2 Question
« Reply #15 on: Mar 08, 2012, 11:55 »
Thank you for your very enlightening replies. I have tried to read any material that is available on this subject and on commercial nuclear power in general. I think it is a safe, clean, efficient way to generate electricity. I think that history has shown that even when things go really really wring that the defense in depth safe guards will protect the public health and safety. I also think that the fact that there have been so few serious incidents is a testament to the training and professionalism of you folks, I am very fortunate to be able to read this site and learn from professionals like yourselves.

On another note I have found the B&W Cross Training Manual (available on the NRC sit)e section on TMI 2 and the Davis Besse loss of feed water event from the 80's to be the most educational and intriguing information out there. If anyone can point to any other good resources it would be appreciated.

mjd

  • Guest
Re: TMI 2 Question
« Reply #16 on: Oct 06, 2014, 11:08 »
This topic is fairly old, however I just read it for the first time and would like to offer some comments relative to less than 100% totally accurate responses. First, the original question has been answered correctly, the "EFW 12s" being initially closed at TMI2 had no effect on the initial primary pressure response, for the reasons stated. There is this comment response "I am not sure if this is correct.  I am betting that the trip and initial SG blow down did not lift the PORV.  The PZR Spray should have been enough to stay away from the PORV lift setpoint.  This is all just an educated guess from me and should in no way be construed as a fact."
The PRZR spray valve, in auto, on a B&W plant can not in fact overcome the RCS pressure increase if the initiating event is a loss of MFW, even from low initial power. This is a fact; during the Sept 24, 77 event (TMI2 precursor) at DBNPP the spray valve did open, the PORV still opened, both before an RPS HP trip occurred. These two actions prevented an RPS HP trip, RCS pressure was decreasing when the RO manually tripped the reactor in response to the PRZR level approaching off-scale high. The observation that a manual 100% open spray valve may prevent PORV lift on the transient may be true, but with the current ARTS system it is about impossible to simulate. The spray valve at TMI2 was in manual at the time of event initiation, for B equilization. When the TMI2 reactor tripped an RO put the spray valve back in auto and hit the start switch for the 2nd MUP, which did not start; it was successfully started slightly later by a different RO. This may have added to control room confusion also.
There is a comment response "Actually in the 70s a turbine/reactor trip on a BW tended to operate the PORVS 100% of the time." This just is not true, I saw plenty of 100% power trips without a PORV lift. It totally depends on the initiating event. The terminology "turbine/reactor" trip is misleading for the '70s/early '80s as a turbine trip did not automatically trip the reactor then, so the terms should not be combined for discussions of that era.
There is a comment response relative to root cause of TMI2 (discussing RCP trip vs HPI shutdown). Much has been written, and officially not written, about the "Root Cause" of the TMI2 accident. The real root cause of the TMI2 accident is identified here: http://www.nukeknews.com/index.html

 


NukeWorker ™ is a registered trademark of NukeWorker.com ™, LLC © 1996-2021 All rights reserved.
All material on this Web Site, including text, photographs, graphics, code and/or software, are protected by international copyright/trademark laws and treaties. Unauthorized use is not permitted. You may not modify, copy, reproduce, republish, upload, post, transmit or distribute, in any manner, the material on this web site or any portion of it. Doing so will result in severe civil and criminal penalties, and will be prosecuted to the maximum extent possible under the law.
Privacy Statement | Terms of Use | Code of Conduct | Spam Policy | Advertising Info | Contact Us | Forum Rules | Password Problem?