Effect of Earthquake on Newer Reactors

[mhs25]

With everything that's going on at Fukushima, I'm wondering how a newer reactor (Generation III or newer) would have held up under the same conditions?

I understand it's near impossible to tell from anything this catastrophic and unprecedented, but perhaps Astronuc or someone else more educated then me (I'm still just a student) could give their opinions.

From what I understand, the ABWR and AP-1000 have passive safety systems, cooling loops that flow by natural convection instead of by pumps. This would make the current problem impossible, yes? What would need to happen in order for this to fail? And would it be more or less likely to fail in a 9.0 Earthquake and Tsunami?

Fukushima Reactor number one was built in 1970 correct? With the others following not long after that. These would be considered Generation II reactors right? Is Generation I just the early research reactors or are there commercial plants that are considered Generation I?

Thanks for any answers you can provide.

 

[Astronuc]

The Gen III+ reactors, the new ones like PWRs: AP-1000, EPR and APWR, and the BWRs: ABWR and ESBWR have or are in the process of being certified.

In terms of natural phenomena, the biggest one is probably earthquakes, which require a certain level of design for seismic loads - accelerations. The beefier or stronger the design, the more places that one could site a plant. Depending on the design, there are limitation on where the plant can be sited. Before a plant is built, the utility is responsible to determine a site according to the suitability of that site.

The may be elements of the designs, particularly the passive cooling reservoirs that may be revisited in order to determine the effect of pool seiches, and/or loads for a given earthquake. A seiche might require more baffles or higher walls, which themselves must resist the forces/loads of whatever accelerations are produced by a given design basis earthquake.

The Fukushima Daiichi, Unit 1 came on-line in 1971, so it was designed in the early to mid 1960's. It is a BWR 3, with Mk I containment.

Many of the Gen I and a few Gen 2 plants have been shutdown and decommissioned. Others are hitting 40 years soon. I'd have to check my library to see if any hit 40 years.

 

[mhs25]

The Gen III+ reactors, the new ones like PWRs: AP-1000, EPR and APWR, and the BWRs: ABWR and ESBWR have or are in the process of being certified.

In terms of natural phenomena, the biggest one is probably earthquakes, which require a certain level of design for seismic loads - accelerations. The beefier or stronger the design, the more places that one could site a plant. Depending on the design, there are limitation on where the plant can be sited. Before a plant is built, the utility is responsible to determine a site according to the suitability of that site.

The may be elements of the designs, particularly the passive cooling reservoirs that may be revisited in order to determine the effect of pool seiches, and/or loads for a given earthquake. A seiche might require more baffles or higher walls, which themselves must resist the forces/loads of whatever accelerations are produced by a given design basis earthquake.

The Fukushima Daiichi, Unit 1 came on-line in 1971, so it was designed in the early to mid 1960's. It is a BWR 3, with Mk I containment.

Many of the Gen I and a few Gen 2 plants have been shutdown and decommissioned. Others are hitting 40 years soon. I'd have to check my library to see if any hit 40 years.

Okay. I have a decent idea about seismic loads, before I switched my major to Nuclear I did a co-op with a company that sells/designs seismic restraints for HVAC systems here in GA. So the NRC reviews each plant site individually right? So each plant isn't an exact cookie cutter design. Can a plant that can withstand earthquakes of a 9.0 scale be designed economically?
Also- I remember hearing that some of the reactors near fault lines in California (Diablo Canyon I believe) are designed to withstand earthquakes of 7.0 (I'm pretty sure that was the number.) Does this mean that they are designed to continue operating even after a 7.0 earthquake, or that systems will start failing like at fukushima with a 7.0+ earthquake?

So do you know if the Fukushima reactors were Generation I or II? Is there really an official definition?

 

[Astronuc]

Okay. I have a decent idea about seismic loads, before I switched my major to Nuclear I did a co-op with a company that sells/designs seismic restraints for HVAC systems here in GA. So the NRC reviews each plant site individually right? So each plant isn't an exact cookie cutter design. Can a plant that can withstand earthquakes of a 9.0 scale be designed economically?

On an economical design for mag 9 accelerations - probably not.

Also- I remember hearing that some of the reactors near fault lines in California (Diablo Canyon I believe) are designed to withstand earthquakes of 7.0 (I'm pretty sure that was the number.) Does this mean that they are designed to continue operating even after a 7.0 earthquake, or that systems will start failing like at fukushima with a 7.0+ earthquake?

I understand that the plant has a challenge with some mag, perhaps 7.0 or greater, but I'm not certain.

So do you know if the Fukushima reactors were Generation I or II? Is there really an official definition?

The Fukushima reactors survived the earthquake - they were shutdown normally per procedure, the shutdown cooling began as expected. I'm not sure at what point the grid went down, i.e. they lost connection to the grid offsite. At that point the emergency diesel generators (EDGs) came on line. THEN - the tsunami took out the fuel supply to the EDGs and apparently damaged some of the electrical work.

IF the EDGs and fuel supply had been on the opposite of the reactor building (and the electrical system as well - or 10 m above the surrounding ground level) - away from the ocean - they could have operated, the reactors would be functioning normally now - and we could be celebrating the fact that our technology survived and successfully operated after one of the biggest earthquakes in history. Instead we are witnessing a failure how we apply technology. Obviously - that has to change.

 

[mhs25]

IF the EDGs and fuel supply had been on the opposite of the reactor building (and the electrical system as well - or 10 m above the surrounding ground level) - away from the ocean - they could have operated, the reactors would be functioning normally now - and we could be celebrating the fact that our technology survived and successfully operated after one of the biggest earthquakes in history. Instead we are witnessing a failure how we apply technology. Obviously - that has to change.

Wow, could it really have been that simple? I knew everything reacted normally to the Earthquake, so defending against the tsunami too can be avoided rather easily in the future?

 

[rogerl]

Pro-nuclear group keeps saying the reactors in Fukushima are old already and newer reactors are much safe. Why. What class of nuclear reactors can survive even if the pumps are without electricity to run coolant to the reactors. Is there such a thing? Even the newer PWR nuclear plant seem to need coolant.

http://en.wikipedia.org/wiki/Pressurized_water_reactor

"PWRs can passively scram the reactor in the event that offsite power is lost to immediately stop the primary nuclear reaction. The control rods are held by electromagnets and fall by gravity when current is lost; full insertion safely shuts down the primary nuclear reaction. However, nuclear reactions of the fission products continue to generate decay heat at initially roughly 7% of full power level, which requires 1 to 3 years of water pumped cooling. If cooling fails during this post-shutdown period, the reactor can still overheat to above 2200 degrees centigrade where separation of water into its constituent elements Hydrogen and Oxygen occurs. In this event there's a high danger of hydrogen explosions, threatening structural damage and/or the exposure of highly radioactive stored fuel rods in the vicinity outside the plant in pools (approx 15 tons of fuel is replenished each year to maintain normal PWR operation)."

 

[Astronuc]

Wow, could it really have been that simple? I knew everything reacted normally to the Earthquake, so defending against the tsunami too can be avoided rather easily in the future?

Yeah - it would have been that simple - i.e., this event was preventable - so was TMI-2 and so was Chernobyl.

 

[Thermalne]

Pro-nuclear group keeps saying the reactors in Fukushima are old already and newer reactors are much safe. Why. What class of nuclear reactors can survive even if the pumps are without electricity to run coolant to the reactors. Is there such a thing? Even the newer PWR nuclear plant seem to need coolant.

http://en.wikipedia.org/wiki/Pressurized_water_reactor

"PWRs can passively scram the reactor in the event that offsite power is lost to immediately stop the primary nuclear reaction. The control rods are held by electromagnets and fall by gravity when current is lost; full insertion safely shuts down the primary nuclear reaction. However, nuclear reactions of the fission products continue to generate decay heat at initially roughly 7% of full power level, which requires 1 to 3 years of water pumped cooling. If cooling fails during this post-shutdown period, the reactor can still overheat to above 2200 degrees centigrade where separation of water into its constituent elements Hydrogen and Oxygen occurs. In this event there's a high danger of hydrogen explosions, threatening structural damage and/or the exposure of highly radioactive stored fuel rods in the vicinity outside the plant in pools (approx 15 tons of fuel is replenished each year to maintain normal PWR operation)."

All reactors need some form of coolant to cool the fuel rods.

 

[Astronuc]

Pro-nuclear group keeps saying the reactors in Fukushima are old already and newer reactors are much safe. Why. What class of nuclear reactors can survive even if the pumps are without electricity to run coolant to the reactors. Is there such a thing? Even the newer PWR nuclear plant seem to need coolant.

The PMBR and gas-cooled, graphite reactor using graphite-SiC clad spherical (pebble) fuel is considered pretty safe. The graphite has a very high metling point.

A former professor once boasted that there could be a loss of coolant accident (LOCA) with an advanced gas reactor (e.g., PMBR), and one could take liesurely walk to a local restaurant, order lunch or dinner with a nice bottle of wine, and take one's time thinking about a solution to the problem of restoring cooling. I thought that a pretty unsettling attitude. Perhaps that was his way of expressing confidence in the concept.

Nuclear fuel performance and predictive analysis requires the capability of multiphysics simulation. It's very challenging and lot of fun.

http://en.wikipedia.org/wiki/Pressurized_water_reactor

"PWRs can passively scram the reactor in the event that offsite power is lost to immediately stop the primary nuclear reaction. The control rods are held by electromagnets and fall by gravity when current is lost; full insertion safely shuts down the primary nuclear reaction. However, nuclear reactions of the fission products continue to generate decay heat at initially roughly 7% of full power level, which requires 1 to 3 years of water pumped cooling. If cooling fails during this post-shutdown period, the reactor can still overheat to above 2200 degrees centigrade where separation of water into its constituent elements Hydrogen and Oxygen occurs. In this event there's a high danger of hydrogen explosions, threatening structural damage and/or the exposure of highly radioactive stored fuel rods in the vicinity outside the plant in pools (approx 15 tons of fuel is replenished each year to maintain normal PWR operation)."

The 2200 degrees centigrade should probably be 2200°F/1200°C, which is a peak cladding temperature (PCT) requirement, and the models generally have those temperatures for very short time - on the order of seconds - not hours or days. This assumes that there is some cooling besides steam - e.g., reflood with water, which implies the emergency core cooling system (ECCS) works.

 

[Thermalne]

The PMBR and gas-cooled, graphite reactor using graphite-SiC clad spherical (pebble) fuel is considered pretty safe. The graphite has a very high metling point.

A former professor once boasted that there could be a loss of coolant accident (LOCA) with an advanced gas reactor (e.g., PMBR), and one could take liesurely walk to a local restaurant, order lunch or dinner with a nice bottle of wine, and take one's time thinking about a solution to the problem of restoring cooling. I thought that a pretty unsettling attitude. Perhaps that was his way of expressing confidence in the concept.

The 2200 degrees centigrade should probably be 2200°F/1200°C, which is a peak cladding temperature (PCT) requirement, and the models generally have those temperatures for very short time - on the order of seconds - not hours or days. This assumes that there is some cooling besides steam - e.g., reflood with water, which implies the emergency core cooling system (ECCS) works.

Do you know which professor stated that?

 

[Astronuc]

Do you know which professor stated that?

I believe he has passed away, but I won't reveal his name.

BTW - I don't believe that PWR article at Wikipedia is rigorously correct. Clearly it is not written by an expert given the errors in some fundamental information.

Reloads or batch sizes (and split batch distributions) are generally plant and cycle specific. There is no typical anymore across the industry for BWRs or PWRs.

 

[Thermalne]

I believe he has passed away, but I won't reveal his name.

BTW - I don't believe that PWR article at Wikipedia is rigorously correct. Clearly it is not written by an expert given the errors in some fundamental information.

Reloads or batch sizes (and split batch distributions) are generally plant and cycle specific. There is no typical anymore across the industry for BWRs or PWRs.

Ah sorry I didn't mean it like that. I'm very interested in Gas cooled and Liquid metal reactors.

 

[promecheng]

Pro-nuclear group keeps saying the reactors in Fukushima are old already and newer reactors are much safe. Why. What class of nuclear reactors can survive even if the pumps are without electricity to run coolant to the reactors. Is there such a thing? Even the newer PWR nuclear plant seem to need coolant.
"

All LWR designs need coolant, both during operation and after shutdown to maintain residual heat removal. To answer your question about what designs can survive without pumps and/or electricity to facilitate cooling, the AP1000 and the ESBWR both are designed with passive cooling features. The AP1000 has multiple gravity fed emergency core cooling systems that are meant to operate passively, i.e. minimal operator interaction and without the need of power. I'm not as familiar with the ESBWR, as it's design has not yet been approved, but I believe it also has multiple passive gravity-fedcooling systems plus natural circulation through the RPV. These are considered Gen III+ designs. In the case of AP1000, Westinghouse is currently building 4 units in China, (Sanmen & Haiyang), and 4 units are ordered here in the U.S. (Vogtle & Summer).

 

[AlephZero]

Yeah - it would have been that simple - i.e., this event was preventable - so was TMI-2 and so was Chernobyl.

The hard part is figuring out what the realistic scenarios will be. After you have done that, designing to meet them often is quite simple, as Astro said.

Clearly somebody decided how high to build the breakwaters etc that were supposed to prevent sea damage on the basis of some criteria. I don't think we have any solid evidence (yet) about whether or not the criteria used were "optimistic", given what was known when the design was made. You can't blame the original designers for not knowing stuff that would be discovered in the future.

I got pitched into seismic design of non-nuclear power plants a few years back, in the context of small (25 ot 50 MW) generators powered by modified jet engines. We know how to design engines that don't fail when aircraft fly through turbulence (and they do have to be designed to withstand that, it doesn't "just happen"). So at the start we thought this should be fairly straightforward, but there was a big problem in trying to figure out what the available data actually meant, in terms of realistic acceleration levels etc.

Of course a 50MW gas-fired electricity generator isn't going to cause a major environmental disaster whatever happens to it, but would be embarrassing (and bad for future sales!) if a "minor" quake turned it into heap of scrap metal.

With aircraft turbulence (and similar things like heavy landings), at least you can put some instrumentation on a test aircaft and go looking for it, with a reasonable chance of finding it in the right weather conditions. The chance of having a seismograph sitting right on top of a quake epicenter is pretty small by comparison.

 

[Astronuc]

Ah sorry I didn't mean it like that. I'm very interested in Gas cooled and Liquid metal reactors.

Ah - OK. There is plenty of information available through INL (Idaho National Lab) and US DOE, Generation IV International Forum (GIF, or Gen-4.org), IAEA and OECD/NEA.


http://nuclear.inl.gov/gen4/
http://www.ne.doe.gov/geniv/neGenIV1.html
http://www.gen-4.org/
http://www.iaea.org/NuclearPower/GCR/
http://www.iaea.org/inisnkm/nkm/aws/htgr/fulltext/gtpcs_1.pdf

http://www.world-nuclear.org/info/inf77.html

 

[Astronuc]

The hard part is figuring out what the realistic scenarios will be. After you have done that, designing to meet them often is quite simple, as Astro said.

Clearly somebody decided how high to build the breakwaters etc that were supposed to prevent sea damage on the basis of some criteria. I don't think we have any solid evidence (yet) about whether or not the criteria used were "optimistic", given what was known when the design was made. You can't blame the original designers for not knowing stuff that would be discovered in the future.

The Fukushima Daiichi FSAR, Chapter 2, Site Characteristics, and Chapter 3, Design of Structures, Components, Equipment, and Systems

Fukushima should have the Japanese equivalent of:

Regulatory Guide 1.070, Standard Format and Content of Safety Analysis Reports for Nuclear Power Plants, LWR Edition

Chapter 2 SITE CHARACTERISTICS

2.4 Hydrologic Engineering
2.4.1 Hydrologic uescription
2.4.2 Floods
2.4.3 Probable Maximum Flood (PMF) on Streams and Rivers
2.4.4 Potential Dam Failures, Seismically Induced
2.4.5 Probable Maximum Surge and Seiche Flooding
2.4.6 Probable Maximum Tsunami Flooding
2.4.7 Ice Effects
2.4.8 Cooling Water Canals and Reservoirs
2.4.9 Channel Diversions
2.4.10 Flooding Protection Requirements
2.4.11 Low Water Considerations
2.4.12 Dispersion, Dilution, and Travel Times of Accidental Releases of Liqufd Effluents in Surface Waters
2.4.13 Groundwater
2.4.14 Technical Specification and Emergency Operation Requirements

2.5 Geology, Seismology, and Geotechnical Engineering
2.5.1 Basic Geologic and Seismic Information
2.5.2 Vibratory Ground Motion
2.5.3 Surface Faulting
2.5.4 Stability of Subsurface Materials and Foundations
2.5.5 Stability of Slopes
2.5.6 Embankments and Dams


Chapter 3 DESIGN OF STRUCTURES, COMPONENTS, EQUIPMENT, AND SYSTEMS

3.4 Water Level (Flood) Design
3.4.1 Flood Protection
3.4.2 Analytical and Test Procedures
3.4.3 NSSS Interface**
3.4.3 BOP Interface**

3.7 Seismic Design
3.7.1 Seismic Input
3.7.2 Seismic System Analysis
3.7.3 Seismic Subsystem Analysis
3.7.4 Seismic Instrumentation
3.7.5 NSSS Interface**
3.7.5 BOP Interface**

**Interface applies to standard designs only. See Appendix A.

There are supporting Reg Guides and NUREGs, particularly NUREG-0800
Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants: LWR Edition (NUREG-0800, Formerly issued as NUREG-75/087)

Somewhere there must be (by law) a justification for the plant and it's components, particuarly safety-related systems or systems that support safety-related systems.

 

[AlephZero]

The Fukushima Daiichi FSAR, Chapter 2, Site Characteristics, and Chapter 3, Design of Structures, Components, Equipment, and Systems

Fukushima should have the Japanese equivalent of:...

Which are probably almost as intelligible to the average non-nuclear-specialist engineer, as the (internationally legally binding) Joint Airworthiness Regulations (JARs) would be to the average non-aviation-specialist engineer.

In the case ot the JARs, they certainly tell you what you need to demonstrate, but they don't necessarily tell you how to demonstrate it. That can be a matter for agreement with the regulators on a particular aircraft or engine type. And (speaking from direct involvement) sometimes "bad stuff happens" which results in a new section being added to the regulations to cover something which nobody had thought about before the bad stuff happened.

These things are all created and interpreted by fallible humans, usually trying their best NOT to be fallible.

 

[AlephZero]

A question of fact - I'm not trying to play political games here.

According to the news media, the reactors themselves were GE designs. Does anybody know what organization(s) has/have the design responsibility for the complete installation, as opposed to just the reactors themselves?

 

[Astronuc]

A question of fact - I'm not trying to play political games here.

According to the news media, the reactors themselves were GE designs. Does anybody know what organization(s) has/have the design responsibility for the complete installation, as opposed to just the reactors themselves?

The utility TEPCO has the ultimate responsibility as the owner/operator and licensee. Really the licensee is ultimately responsible, which is the case in the US.

TEPCO would have hired an Architect and Engineering (AE) firm.

I believe the AE for Unit 1 was Ebasco, and they possibly did or contributed to the other units, or at least the plant site.

GE is responsible for the NSSS (nuclear steam supply system), the AE does the design of the mechanical and civil works, then there is the construction company, and subcontractors. However, all this work is ultimately under the supervision of TEPCO, which is responsible for the quality assurance of the work and then the operation of the plant.

Now Ebasco is long gone. They got absorbed into Raytheon Engineers & Constructors (formerly United Engineers & Constructors). REC got bought by Washington Group International. WGI acquired Morrison-Knudsen Co.

http://en.wikipedia.org/wiki/Ebasco
http://en.wikipedia.org/wiki/United_Engineers_and_Constructors

http://en.wikipedia.org/wiki/Washington_Group_International

WGI was acquired by URS.
http://en.wikipedia.org/wiki/Washington_Group_International#Acquisition_by_URS

Once upon a time, there was also a General Electric Technical Services Company (GETSCO), but I believe they no longer exist. Perhaps parts were sold, and other parts were absorbed into GE.

Now there is GE Energy, and GE Hitachi (GEH, US) and Hitachi GE (HGE, Japan) to sell the ABWR and ESBWR.

 

[joema]

...Does anybody know what organization(s) has/have the design responsibility for the complete installation, as opposed to just the reactors themselves?

I'm sure this will be meticulously scrutinized in the weeks and months ahead.

I believe for similar reactors in the U.S., the diesel fuel tanks are stored underground. In hindsight putting the tanks on the seaward side of an earthquake-prone area seems a poor idea. Even the term "tsunami" is of Japanese origin.

The first order of business is obviously the immediate disaster, but these other items will eventually be discussed both formally and informally.

 

[MadRocketSci2]

I seriously hope at the end of this we learn lessons and ensure everything that went wrong at Fukushima cannot go wrong at newer and existing plants. I hope we don't instead use this as an opportunity to witchhunt every retired engineer and his grandson that was involved in the design of these plants. There are many things you simply cannot forsee before nature kicks you in the ***. The intelligent thing to do would be to fix the problems, not punish people.

I say this because I'm aware of a few engineering witchhunts in the past - for things that in some cases weren't even the fault of the designers. (Operating aircraft decades beyond their design life, then suing retired designers that worked for companies that don't even exist anymore). There always has to be someone to blame and punish, and it doesn't contribute to fixing anything.

 

[joema]

I seriously hope at the end of this we learn lessons and ensure everything that went wrong at Fukushima cannot go wrong at newer and existing plants...

Many existing U.S. reactors are very similar to Fukushima, but the specific thing that failed at Fukushima was the diesel generators, due to tank location. To my knowledge that wouldn't happen at U.S. plants (tanks are underground), but other conceivable failure sequences might.

If the goal is a passively safe reactor that can thermally sustain long term total power outage, that takes a totally different Gen III+ design. It's an intrinsic design element -- you can't retrofit that. To my knowledge none of the 450+ power reactors in the world use that, as it's still being certified.

Likewise the spent fuel pools are a long-known weak area. Even if the reactor itself handles a sustained total "station outage" through passive safety features, it makes little difference if the same outage causes adjacent spent fuel pools to melt, catch fire and contaminate the environment.

So a total solution is needed, not just a reactor-centric solution.

The good news is techologies exist for such a total solution which address passive safety, spent fuel, waste problem, etc. Whether any of these will be deployed is unknown.

 

[minerva]

The PMBR and gas-cooled, graphite reactor using graphite-SiC clad spherical (pebble) fuel is considered pretty safe. The graphite has a very high metling point.

A former professor once boasted that there could be a loss of coolant accident (LOCA) with an advanced gas reactor (e.g., PMBR), and one could take liesurely walk to a local restaurant, order lunch or dinner with a nice bottle of wine, and take one's time thinking about a solution to the problem of restoring cooling. I thought that a pretty unsettling attitude. Perhaps that was his way of expressing confidence in the concept.

They actually tested that... with the German AVR pebble-bed prototype. Well, they didn't actually test LOCA, but they tested a loss of coolant flow. Something similar was done with EBR-II as well, and they both remained in a stable state, passively, without coolant flow.