Bourne Again Shell (Thanks 15-123!)

A lot of nuclear engineering (power reactors) seems to be solving the Boltzmann transport equation with the diffusion approximation.  This involves taking the cross sections over all energies (essentially 10,000 or so energy groups) into 2-4 energy groups - basically just taking a weighted average.  This is done (mostly) using a program called SCALE.

As temperature changes the material composition changes, and then the multi-group cross sections have to be updated.  This has been done historically by just running SCALE at a bunch of different temperatures.  You know what the idiots in my group wanted to do?  They wanted to run all of the inputs manually.  In less than 15 minutes I created a bash script to run all of the inputs for us, and even parse the relevant output files.  Showed the "I think I'm a real programmer, I use FORTRAN" dude what's up.

And for the most part we don't observe much of a trend in the expected output - they are not supposed to be flat; so the SCALE input deck is probably wrong.

Wave Reactors

I am pleased to announce that UT's ANS contingent beat the combined Michigan / MIT / Illinois ANS team with two wins in pool volleyball.

Looking back I don't think that I did a very good job at explaining what I meant in yesterday's post about the AP1000.  The point I wanted to make is that the power industry seems to be impossible to innovate / enter without a large amount of capital, and because of the regulation mess that can be the nuclear industry it is that much harder.

But there is hope, my friends!  The wave reactor (promising design by TerraPower)!  Basically the idea is to only have one fuel enrichment, and then breed the rest of the fuel during the reactor lifetime. Like a perpetual wave . . .?  (Plotting the k's (ratio of neturons in current generation over previous generation, k=1 is sustained one to one chain reaction) versus time does kinda look like a wave).   A design being perused by a joint Michigan / TerraPower venture is a molten salt design that has some problems, namely needing to pump the salt at 10-15 m/s (it's like pumping chocolate syrup at the 1.2 times the estimated average cursing speed of an unladen European swallow - hows that for an analogy?). 

There are certain design aspects I would change.  I would get rid of the fuel assemblies - they add unnecessary material which slows down the neutrons, but might be need to contain the fertile material in close proximity to the fissile material.  I would dissolve the fuel into the coolant and then use chemistry to extract out the nasties (and actually you can make quite a bit of money selling some of the decay products).  This way it could then be online refueling (currently you have to shutdown LWR for about a month every 1.5 to 2 years to refuel, the only online refueling are the CANDU reactors of Canada) without the problems of pebbles (they might crack).

Tomorrow is the last day at the conference, and then it is off to the Everglades to go Alligator Wrassling.  I kinda want to find a small one and bring it home so Eli has a pet; mostly just to see the look on my Mother's face.  We could just fence it a small section of a turkey pen - that way feeding would take care of itself!

AP1000

The AP 1000 is the reactor design by Westinghouse that is being / might be built in the US (The two utilities have made holes in the ground, but they haven't gotten their Construction and  Operating License yet, so they cannot build any nuclear related components).  In China they are farther along (different regulations and whatnot).


The AP1000 has been projected at the ANS meeting as the future of nuclear power - and some of it is deserved as being the first attempts at new builds in the US (not including finishing up Watts Bar Unit 2).  Look, however, at the callout number 4.    The idea for this aspect of the passive system is that in a reactor emergency the water flows down the pressure vessel, setting up convection currents in the vessel and radiating heat by evaporation from the vessel. system for reactor cooling relies on a water pool above the pressure vessel.   One of the more obvious lessons from Fukushima was that putting water above grade for essential functions is a bad idea.

I am sure that Westinghouse / the NRC has noticed this, and are probably working on deciding if this is a real issue (it is not the same thing as Fukushima, at Fukushima the spent fuel which produces heat was above grade, and then there was cladding failure issues when the spent fuel become drained) and if it is fixing it.  (I mean, we gotta have earthquake proof containers by now - just make a nalgene tub)

The nuclear industry seems to be fixated on the technology from 60 years ago - water cooled reactors.  It is very silly; the licensing requirements are such that it takes a tremendous amount of money to (billions) to certify new designs, and neither the national labs or industry want to spend it (but we can spend 15 billion or so on Yucca Mountain).  If you think about the innovative energy start ups they are on the scale of millions.  Why can't nuclear power be like that?  Sure the are companies out their trying (mPower, for example) but the system is such that it will be very difficult for them to ever get their design licensed.

It is almost like the utilities and NRC got together and collaborated on best how to retard growth in the industry.

Sorry for the negative attitude; I really do believe that nuclear energy has a role to play.  Just slightly disillusioned with the system.

Blue Box (MSR)

Blue Box was a simple reactor design that I did using one group neutron diffusion in MATLAB.

I decided to skip a water moderated reactor because of their low thermal efficiencies and steam generator issues and go directly to a Molten Salt Reactor, based, of course, on the the thorium fuel cycle.

I found that the density of a LiF based molten salt is around 1.8 g/mL.  I used a salt composition similar to the Thorium Molten Salt Reactor in Table 3, but couldn't find a fuel composition so I made up my own:  80% LiF, 10% (HN)F4, 7.5% Th232, and 2.5% U235.  (Th is a not fissile but it is fertile - so I need a fissile material, U235, to breed more fissile materials from the fertile Th232).

The graphics are somewhat terrible, but the reactor core is a box 68 x 68 x 18 cm surrounded by a 1 cm thick SS-316 shield (stainless steel - this can be improved to be a better reflector material), which in turn is sourrounded by 5 cm of graphite, which true nuclear engineering fassion in turn is surrounded by 10 cm of concrete.

After running 110 criticality calculations my final k value is 0.4763 with a standard deviation of 0.0053.  If k is equal to one than there is a chain reaction, so my k being less than one means that the fission won't be sustained; I have a sub-critical reactor.  If you look at the fuel neutron cross sections, you see that the total cross section (dashed line -1 on the legend) is about 1-2 orders of magnitude above the fission cross section for the high energy range (1-10 MeV or so) where the reactor is operating.  This means that there is about a 10% to 1% chance that a neutron will fission as opposed to scattering into a lower energy or being adsorbed.  I can either play with the fuel by adding more fissile materials or try different geometries in order to get a higher k value.

ANS Conference (Sunday)

I said y'all last night.  For shame, for shame!

A large amount of time was spent on Sunday on Fukushima.  I don't know how I feel about that - I feel that the more time the nuclear community spends on it gives the impression that the event was more serious than it is.  Kinda like how brooding on something makes it appear worse.  (Not saying that it wasn't serious and what not).

Some interesting tidbits of information:

• The terabequral exists! (A bequerel is a decay / sec, but I have never heard anybody use the 'tera' prefix in science without talking about floating point operations.
• A guide to the crisis level.  I still think this is a lot like the terror alert system.
• A short quiz on 'clean' energy from AREVA (I am think about working for them next summer between my masters and Ph.D.)  The quiz itself is kinda dumb, but the high score names are funny - they are names of reactor designs and famous reactors.

Hydrogen Production!

I pretty much stole the chart from Yildiz, Kazimi 2006.  I don't know much about the benefits of hydrogen fuel cells over better batteries, but I like the idea of developing new types of reactor.  You can read the paper for more details on each method.  I included Steam Methane reforming as an option because they did (and it is the most common commercial method), but as it uses methane and produces carbon dioxide I don't think that it will be a good hydrogen production method in the long run.  Of the reactor designs I like the LMFR (liquid metal fast reactors) the best because they close the fuel cycle and I think have the best passive safety because of their pool type design. They have been built before - The Russians use them on their Alpha Class submarines.

Pool Type Design (Pool provides a large heat sink)
The Soviets used a molten lead fuel for the Alpha Subs, while the above is Sodium

On an unrelated note, some poor bloke at Syracuse has the user ID 867-5309.  You can use it to get a copy of Office 2010. Try it - I tried it as a joke (long story) and was amazed.

Steam Generator Acoustic NDE

Some preliminary results from the steam generator acoustic non-destruction evaluation.*  In the below charts, the difference between the original pipe and that pipe with the noted defects are plotted.

Crack Defect

Hole Defect

We have proven that there is significant difference in the frequency response, and that a hole causes a noisy spectrum than a crack. The axial crack was quite long; it was a hack saw cut into the pipe, and for a real steam generator if you had a 5/64" hole in your steam generator there would be a huge problem - radioactive water would be mixing with the turbine steam.

*The idea is to detect defects in a steam generator by essential vibrating the entire structure. The principle is that if you ding a pipe with a long crack compared to one without you can hear the difference.

Japan's Nuclear

A link that was sent out to our department.

U.S. NRC on the Future of Nuclear Energy

Our colloquim today was on the past and future of US nuclear energy.  It was given by William Magwood (one of the five commissioners of the NRC, coiner of the term Gen-IV), who got his BS in physics from CMU.  Naturally we did the secret physics hand shake and I now have a high paying job in the NRC.   No, but I did introduce myself to ask if Garoff had a comb over back then.  He didn't remember; I call shenegians.

His talk was kinda depressing.  There is only 1 plant in the US that is currently doing actually construction; and that is Watt's Bar Unit 2 which doesn't really count because TVA is just completing the work it stopped (Watts Bar 2 went from being 92% completed to ~70% completed - TVA sold off parts).  On slide 5 he showed the nuclear reactors under construction and planned in the world (from this data) and it is surprising how few they are; and how America is lagging behind.

The plants that are currently under construction are Gen III+, and none of them have been licensed for use in the US yet (although two should be done within a year).  There are currently two small module reactors under consideration for licensing - NuScale and mPower (both are water cooled, mPower is being considered for Watt's Bar 3&4, and Bellefonte has plans to install 6 SMRs, I think mPower as well).  It seemed that all of these changes are going to be happening just in the future - always.

The worst was when he was asked about newer reactor technologies (Molten Salt, Gas Cooled (which I am interested in)) he matter of factly said that the US would probably be importing that technology.  Looks like I got to learn French.

Nuclear Reactor Simulation

I found an online simulation game of a nuclear reactor.  On first play I caused a meltdown in 7 days.  Can you beat that?

BlueBox

Introduction
BlueBox is a homemade cooler at Pappy's Orchard that has 6" thick extruded Polystyrene walls.  Polystyrene is a good neutron moderator and reflector, so the idea is to put a small reactor into BlueBox in order to provide power for Pappy's Orchard and Pappy's Orchard, turkey division.

Simulation

Core Geometry

I did a basic simulation of a 1x1 reactor with nine water inlets.  The core was just U-235, and for both the water and the core I used single group diffusion; i.e. thermal averages of the adsorption and fission coefficients were used.

I had to do an simpler similar problem for a course project, but in that project we had to solve the diffusion equation with a linear system of equations; for this I used MATLAB's PDE toolbox (it made the more complex geometry easier to solve and is a lot faster).

Core Flux

The core flux of neutrons is on the left.  Water is a moderator (slows down and absorbs neutrons), so in and around the water inlets the flux drops.  You can also see the general cosine shape of the flux, the analytical spatial solution for a homogeneous reactor (no water inlets) is

$\Phi(x,y,t)=\sum_{n~odd}^{\infty}e^{-\lambda_n t}\left ( C_n Cos(\frac{\pi x}{n a})Cos(\frac{\pi y}{n a})+ B_n Sin(\frac{\pi x}{n a})Sin(\frac{\pi y}{n a}) \right )$

Here the sin terms drop out because I required the flux to go to zero at the boundary, resulting in:

$\Phi(x,y,t)=\sum_{n~odd}^{\infty}e^{-\lambda_n t}\left ( C_n Cos(\frac{\pi x}{n a})Cos(\frac{\pi y}{n a})\right )$

Thorium Liquid Metal Reactor

Reposting an article that summarized very nicely the thorium and liquid metal reactors. I am very interested in these designs.

The article can be found as the first link in paragraph one.

I find it interesting that in our NE colloquium the presenters talk about improving Light Water Reactor designs - we have yet to have one on fast breeders, liquid metal, or other designs.

Passive Safety Systems

Passive safety systems are based on natural forces (convection, gravity) and do not need any external input (operator / energy) to work.  I think the idea is really great - why design complex "active" systems when you could let nature do what it does?  However, in reading reviews about passive safety systems I find that most are constructed as backups to the active systems.  For example, the EPP-1000, ESBWR, CANDU 6 & 9, AC-600 and AP-600 (among others) all have a large water tank over the reactor. The design is that after the active system fails, gravity just dumps water all over the reactor, cooling it.  Another common design is to build a large "containment pool" that collects the reactor after it melts.

I don't understand why the passive systems are being used as a last attempt backup instead of the first.  It would just seem more natural to use a system that works off of nature instead of pumps and machines to prevent accidents.

There are problems with passive systems.  They tend to be slower to respond / alleviate the problem (gravity driven and heat transfer can be much faster with pumps), and it they might not work outside a given problem / setup, while active ones have more flexibility.  (You can always pump fluid, but with a passive system you need something to drive the flow).