Cold shutdown that doesn't require coolant circulation?

Re: "cold shutdown" that doesn't require coolant circulation?

You mean design a reactor so the fuel can be completely removed?
What makes you think this does not already happen?

 

Re: "cold shutdown" that doesn't require coolant circulation?

Fukushima Unit 1 had such a system - Isolation Condenser. It is nearly passive - no pumps required. The only active measurements needed to operate IC for days is to pour more water into IC tanks as it evaporates.

Nothing prevents IC designs with larger tanks and/or with more trains (Fukushima had 2). So yes, fully passive cooling is possible.

As far as I understand, other units don't have IC for a quite appalling reason: for units with bigger design power, IC also needs to be bigger. Active systems can be made more compact. So IC was replaced by them.

There goes "safety first" PR of nuclear industry... in practice, they did sacrifice safer systems for somewhat more compact active ones.

 

Re: "cold shutdown" that doesn't require coolant circulation?

The AP1000 westinghouse design still requires circulation, but it uses natural forces, and can keep the reactor cooled for 72 hours with no human interaction or electrical power once the initiation signal fires off. Gen 4 designs are primarily passively cooled and require no active systems at all for extended periods of time (as low as a week or indefiniately)

Regarding nikkom's comment about the IC being removed, the IC is a great system in concept, but for the evolution of the BWR it was a bottleneck. The IC in BWR series plants only contains enough water for about 20-30 minutes per train (at 2 trains is about an hour at most). IC was replaced by RCIC. RCIC is a pump which is driven using steam from the reactor. It initially draws water from an outside tank and can function for up to 8 hours using external water before switching over to the suppression pool for cooling. It requires no AC power, only DC power to initiate and control. It can also be black started without DC power, but it requires manual control.

RCIC has a much longer injection and cooling time than IC does. Additionally IC ONLY removes decay heat, it does not provide any injection. RCIC is capable of removing a fair amount of decay heat AND injecting.

The RCIC system at Fukushima unit 2 ran for 70 hours. Unit 3 ran on RCIC for about 32 hours (not sure why it failed, probably loss of DC power coupled with no injection water to use and overheated suppression pool water flashing in the pump causing pump damage....but we wont know for a while. Without active cooling or an external water source, RCIC with no DC power will eventually overheat and fail, but it still lasts for drastically longer than the IC would have). If unit 1 had a RCIC system, the whole accident may have been less of an issue. The operators wouldnt have wasted so much time on trying to figure out if the IC was functioning or not and would have been able to prioritize real issues. They spent a disproportionate amount of time and resources trying to get the IC to work in unit 1. Personally I think IC plants should be shut down. RCIC all in all is the better system. IC requires an active feed pump to maintain it filled, while RCIC self powers and pumps using reactor steam. The plant I'm at was notorious for running for several days on RCIC after BOP scrams to maintain the reactor in hot standby and save a few hours off of their next restart. IC also contaminates the area outside the plant. It's low level, but the IC uses reactor grade water on the shell side, which contains small amounts of tritium and some other radioisotopes. Dresden plant in illinois (before security fences prevented this), they would have to decon the first 2 or 3 rows of cars outside the reactor building after running the IC.

 

Re: "cold shutdown" that doesn't require coolant circulation?

In Fukushima Unit 1, IC capacity was 8 hours of cooling. It's trivial to design an IC with bigger water tank.

Pump is an active system. It can break, it can't function if steam pressure is lost (say, a fissure in RPV's top).

That's exactly why we want to see passive systems, which don't need DC to work.

IC tank is at atmospheric pressure and can be refilled by a very ordinary equipment. A fire truck will do. Try using it to cool overheating suppression chamber. Good luck.

And if operators would have even minimal training for SBO, and in particular, how to activate the IC during SBO, the disaster may be averted altogether.

???!!!

WHAT??? Why a clear water can't be used in IC?

 

Re: "cold shutdown" that doesn't require coolant circulation?

I just pulled the FSAR for a plant which contains an IC [Dresden http://pbadupws.nrc.gov/docs/ML0719/ML071910096.pdf Same model as unit 1. Go to section 5.4.6]. The IC is made up of 304 SS tubes, and contains enough water inventory for about 20 minutes prior to boil-off. The IC's accredited makeup source is the condensate storage tank (CST), via diesel driven transfer pumps. The condensate transfer pumps, fire protection system, and probably some other stuff, can be used to get water into the IC. The valves for the IC still require AC power for the MOV and control power to be actuated (so if they were potentially closed, like when the tsunami hit unit 1, it's useless). I'm not sure if they fail close on loss of power.

When you look at the water in the CST, there is more than enough for 8+ hours of IC for decay heat removal. But the shell of the heat exchanger only has 20 minutes of water. (If you want more evidence of this, go to the NRC teleconferences the day of Fukushima. They are available through the FOIA links. They state it in there, each IC has 20-30 minutes of water in it. I think its in the first of like the 7 teleconferences posted.)

>Pump is an active system. It can break, it can't function if steam pressure is lost (say, a fissure in RPV's top).

RCIC and IC are not ECCS systems, and are not required to function for a primary system line break. The RCIC pump is powered passively (decay steam) which gives it some inherent advantages to something which fails after 20 minutes. If you had a line break, HPCI(high pressure coolant injection)/HPCS(high pressure core spray) are accredited for high pressure injection. Assuming single failure of HPCI or HPCS, then on level 1 water level (low-low-low alarm water level), the ADS (automatic depressurization system) activates to blowdown the vessel and inject with all three LPCI (low pressure coolant injection) systems and the LPCS (low pressure corespray) system. Neither RCIC nor IC are accredited for loss of primary loop integrity accidents.

>That's exactly why we want to see passive systems, which don't need DC to work.

Additionally, RCIC's main limit is suppression pool temperature, as it uses that water to cool itself when it is in recirculation mode. During SBO, you lose RHR's ability to remove containment heat, and containment venting is the only heat removal you have to attempt to reduce suppression pool temperatures.

>IC tank is at atmospheric pressure and can be refilled by a very ordinary equipment. A fire truck will do. Try using it to cool overheating suppression chamber. Good luck.

IC doesn't inject water into the vessel or containment. It also does not cool containment. RCIC tank is at atmospheric pressure and can be refilled just as easily. Additionally RCIC injects, removes decay heat, and can help manage suppression pool and containment temperature/pressure (adding colder water into the vessel from the outside rather than recirculating suppression pool water).

>And if operators would have even minimal training for SBO, and in particular, how to activate the IC during SBO, the disaster may be averted altogether.

One of the identified issues (as can be seen in INPO IER 11-05 August 2012 addendum found here: http://www.nei.org/resourcesandstat...t-the-fukushima-daiichi-nuclear-power-station), is that Japan deviated from many decisions and lessons learned the US had after Three Mile Island. One of those lessons learned was that operators should train in a simulator that matches the plant they are working in. At Fukushima, operators trained in the Unit 2-4 simulator, as they did not have a unit 1 simulator, and as a result never had "hands on" experience with the IC. That lack of knowledge combined with no AC/DC instrument power or control power made figuring out if unit 1 had cooling near impossible. If unit 1 had a RCIC system, this wouldn't have been an issue, as RCIC does not need to be cycled to prevent violation of reactor vessel cooldown rate technical specifications like the IC does, and would have remained in service as it did in unit 2 and 3 when the tsunami hit.

>WHAT??? Why a clear water can't be used in IC?

Editing this section: Clean water or reactor grade condensate can be used for the IC. Clean water is much more expensive to make, and may not have been the preferred source 20+ years ago (which is where the stories I know of regarding contamination come from). Condensate contains low level contamination. The IC system does have radiation monitors which will isolate the atmospheric vents at high radiation levels, but this is intended to protect against a leak path from damaged fuel, out through a damaged heat exchanger tube, to the atmosphere. I do not believe there is filtering on the atmospheric vents.


Personally I think the IC is a big danger. It requires active diesel pumps for more than 20 minutes, and you still need control power to line it up. It can't inject (so you are still reliant on HPCI if you are in hot-standby and need to inject), and it cools the vessel down faster than the 100 degree F per hour limit on BWR reactor vessels and requires operators to cycle it constantly. With RCIC, it will auto stop and start between reactor water level 8 (high high alarm) and level 2 (low low alarm) [respectively]. It has a flow controller and can be dialed to inject to match decay heat directly to prevent excessive cooldown, while injecting. It can inject outside water. It functions with no AC, and its pump does not require an active energy source (as decay heat is the energy source). Maybe I'm slightly biased, but I restate my point, at Fukushima, unit 1 was a huge concern, and it had no RCIC, while units 2 and 3 lasted 70 and 36 hours respectively.

Side note: Great discussion/comments.

 

Re: "cold shutdown" that doesn't require coolant circulation?

As nikkkom said, IC capacity of the order of 8 hours has been built e.g. at Fukushima Dai-ichi unit 1, as well as several other plants of the same generation. 20 minutes is obviously too short, but it doesn't mean the IC wouldn't - if designed properly wrt. capacity and valve fail safe modes - be a great device for passive shutdown cooling.

 

Re: "cold shutdown" that doesn't require coolant circulation?

I'm not sure if I understand, but again, the IC does not have 8 hours of capacity built in, it has 20 minutes, and requires active systems to function for longer than that.

The IC as it exists in BWR/2 and /3 plants is pretty awful as it is limited capacity and requires active pumping systems, where RCIC can run for DAYS.

Now in comparison, the IC as designed in the ESBWR, and the PXS heat exchanger in the AP1000 (basically an IC), are excellent, but that requires a full plant/containment design to support it. The IC as exists in older BWRs is a vulnerability in my professional opinion, and I thats what we saw at Fukushima.

 

Re: "cold shutdown" that doesn't require coolant circulation?

You might be right about that particular design. I don't know the details.

If operators would engage IC on Unit 1 and it didn't save the day, I'd agree with you.

But that's not what happened.

In Fukushima after SBO, operators failed to keep valves from RPV to IC open and thus IC couldn't do its job.

 

Re: "cold shutdown" that doesn't require coolant circulation?

There appears to be different designs of the IC regarding the shell size capacity. that post by Tsutsuji-san cites the Japanese NSC saying

The IC was run for an hour between the quake and tsunami, and its failure was due to spurious isolation signal due to loss of DC, as translated by Tsutsuji-san in the "Japan Earthquake: nuclear plants" thread around page 750. Last October, when the IC:s were checked, they were still more than half full, so running out of shell side water was not the cause of loss of the IC capacity - it was the spurious isolation signal.

 

Re: "cold shutdown" that doesn't require coolant circulation?

They manually closed the valves prior to the tsunami hitting (actually this happened several times, they cycled the IC in and out of service because the IC cools the reactor down too quickly). But yea, once they lost control power and instrument air it was pretty much over for unit 1. BWR vessels have a 100 degree F/hr tech. spec. cooldown/heatup rate, and each reactor vessel is only rated to exceed this once in its lifetime. To date, I do not believe any plant has actually done so (at least not in the US). Based on differing reports, they were trying to re-start the IC, and its not entirely clear what happened. I agree with the above poster than an isolation signal may have come in on loss of DC, but at the same time, the AC system was out of service around the same time, so it is possible the valves were half shut or in some unknown intermediate state. Regardless they spend a LOT of time trying to figure out the state of the system which could have been better spent on other issues.

If the valves were lined up correctly and the diesel driven pumps were functioning and the makeup tank was still intact then yes, it would have helped a LOT at Fukushima. They still would have been limited to 8 hours, but that would have let them prioritize getting DC power and instrumentation restored.

 

Re: "cold shutdown" that doesn't require coolant circulation?

Just posting another follow up, I likely made an error. While Dresden is nearly identical in design to Fukushima Daiichi, Dresden operates 2 reactors at over 800 MWe (over 2400 MWth), while Fukushima Daiichi unit 1 is a single unit which operates around 460 MWe (just over 1380 MWth most likely).

This difference in thermal output also directly correlates to a difference in decay heat, and would explain why Dresden only accredits their IC for 20 minutes while Fukushima Daiichi news posts claim up to 6 hours on shell side water inventory.

My opinion is still the same regarding RCIC vs. IC, but I apologize if I made this confusing.

 

Re: "cold shutdown" that doesn't require coolant circulation?

I think he may have misphrased this one. An IC doesn't necessarily contaminate the area outside the plant, but in case of leaks there is the possibility of it doing serious contamination.

As far as I understand both systems, RCIC is fully integrated into the PCV, therefore any pipe leak anywhere in the piping would vent coolant into the primary containment, but not into the environment.

Contrary to that, the IC-piping naturally has to be at least partially outside the primary containment - since the coolant pool is there as well. Therefore in a reactor with an operating Isolation Condenser, there's a direct and unfiltered connection between the RPV and the environment, with the primary containment vessel being circumvented. If the IC piping outside the PCV would be damaged, the fuel would be directly connected to the atmosphere.
And that's the reason why it had been constructed fail-safe. You absolutely don't want a backdoor in your containment you may not be able to close in case of accidents. Therefore make sure that it's closed on default.

To go back to Fukushima:
I'm not so sure anymore if constructing it fail-safe had been the wrong decision. Let's say it had been constructed fail-operational.
Then Unit 1 would have survived for 8 hours with adequate cooling, after that the reactor would've boiled dry anyway, high pressure steam would have ruptured the IC-pipes outside the reactor and fission products released from the failing fuel rods would've found their way directly into the atmosphere.

 

Re: "cold shutdown" that doesn't require coolant circulation?

The IC shell side can be filled using condensate water (primary loop), which IS contaminated. Dresden had to regularly decontaminate cars in the parking lot after IC use in the early 90s before they switched to using demineralized water instead.

The IC is not a direct unfiltered path, as the heat exchanger tubes are the RCS boundary in this case. The IC does have rad monitors which will isolate the IC should high radiation levels which could indicate an unfiltered release are detected, but that shouldn't be happening as the heat exchanger tubes shouldnt have leakage. The isolation system valves fail closed to ensure that if the IC would become part of the RCS, that there would not be a release path.

Also the IC pipes should not rupture. They are ASME code, so they should be tested to withstand 3 times reactor pressure. Additionally, the safety relief valves prevent the RCS from exceeding 1300 PSI, so the IC would never have to deal w/pressures that high (in fact, the IC, even if it was open at unit 1, would never have had pressures above 1200 due to SRV lifts)

 

Re: "cold shutdown" that doesn't require coolant circulation?

Well, some time ago I found a patent of a (newer) isolation condenser where they specifically addressed the problem of bursting pipes:

https://www.physicsforums.com/showpost.php?p=3881754&postcount=12998

 

Re: "cold shutdown" that doesn't require coolant circulation?

For nuclear reactor safety analysis you HAVE to assume you are going to have pipes burst. It is a requirement. That's precisely why isolation valves exist. The pipes are ASME code and shouldn't burst or rupture, but you need to have defense in depth to get licensed to operate a nuclear power plant. It is for that reason that the IC valves are energize to open, and its also why there are radiation monitors which will close those valves if high radiation levels are detected leaving the plant.

This is consistent with GE designed isolation valves in other systems, including the main steam lines, the reactor building normal ventilation system, drywell purge systems, etc. When high radiation is detected in a system, that system is isolated.

The patent doesn't add anything new about pipe burst. A pipe burst can always happen. But the key thing here is it shouldn't.

It's just like LOCA. All the reactor piping is designed, inspected, and maintained in accordance with ASME codes and legal requirements, and as such, the piping should NEVER break in the lifetime of the plant. However you still have to assume they do, in order to overdesign your facility and ensure that you can prevent a release of radioactive material.

 

Re: "cold shutdown" that doesn't require coolant circulation?

In later ASEA BWRs, the containment isolation valves in the emergency cooling lines do not fail close, and their isolation signal is generated by energizing the circuit, unlike in other penetrations. I don't know if that (preferring core cooling to containment isolation) would be possible under the US regulations.

 

Re: "cold shutdown" that doesn't require coolant circulation?

It really depends. In the US you have to assume certain types of failures then calculate the maximum leakage from the containment under those situations to come up with a dose delivered to the public. Any valves which don't fail closed would probably need to be considered open, and it would just count as a penalty against your accident analysis.

There are some valves that require energy to close, but typically they aren't main steam valves.

Thanks for the interesting info though. I wouldn't have thought ASEA would have gone in that direction. I know GE BWRs over seas do some stuff that the US BWRs would never dream of doing (island startup, REVABS, RHR steam condensing, etc).