Reactor design/systems design

[aliaze1]

I am currently a junior (senior in a week haha) in Electrical Engineering, concentrating in Power/Control Systems.

I took an introductory course in Nuclear Engineering and have become very interested in reactor design/system design (BWR/PWR/Submarine reactors/entire systems not just the core).

I plan to go to graduate school for NE, but what level of education would one need to work for a company that designs reactors/systems (such as Westinghouse)? a few people said that at that level of work I may need a PhD, which I don't mind going for, but I figured I'd ask. Also, is there anything aside from control systems that I should take as an EE that would be beneficial?

 

[TVP45]

Many of the senior engineers and managers I have met from Westinghouse are either PhD or ABD. It really depends on what part of the design you want to do. I would recommend an MS an a minimum. But, they have an aggressive hiring campaign on this spring; why not contact them and ask what they're looking for?

 

[aliaze1]

Many of the senior engineers and managers I have met from Westinghouse are either PhD or ABD. It really depends on what part of the design you want to do.

hmm I am not 100% sure yet. someone had to come up with the CANDU idea, or the PWR idea, or any of the various submarine/boat ideas. that type of position...if i were to be more specific i guess i'd do reactor safety/system safety stuff

why not contact them and ask what they're looking for?

good idea, I should check out their website

 

[aliaze1]

either PhD or ABD...an MS an a minimum

you're right on the dot

"[URL [Broken] Advanced Reactor Designer[/URL]

"MS in Nuclear Engineering minimum, PhD preferred."

 

[Astronuc]

Vendors would prefer someone with at least an MS. A PhD might be preferable, but it's not necessary.

One would need courses in neutronics (core design, core simulation, reactivity control), structural materials (familiarity with pressure vessel design and mechanics of PVs), and thermal hydraulics (prefereably with some CFD experience).

 

[aliaze1]

I have taken Statics and am planning on taking Strength of Materials this summer.

The remaining semesters I have almost all EE courses

[Power systems (2 classes), Electric Machinery (1), Electromagnetics (2), Electronics (1 lecture, 1 lab), Probability/Statistics for EE's (1), Control Systems (1), C programming (1), Logic design and Digital systems (1 each)] I might be able to squeeze in a Plasma/Fusion science course in one of the semesters

the non EE courses are [Quantum Mechanics 1, Strength of Materials, possibly a Reactor System Safety Analysis course if I can get it approved for an elective, taken intro to NE and Statics, can also take dynamics if deemed necessary]

The Neutronics courses (as well as other nuclear courses) would be taken in graduate school

I have registered for a quantum mechanics course, would that be useful?

As for thermal hydrualics, I do not know exactly what this subject is, but someone did suggest I take fluid mechanics and thermodynamics (though i despise thermo...but that might be because of the professor i had)

 

[Morbius]

As for thermal hydrualics, I do not know exactly what this subject is, but someone did suggest I take fluid mechanics and thermodynamics (though i despise thermo...but that might be because of the professor i had)

aliaze1,

"Thermal-Hydraulics" is the term for the coupled physics problem of fluid mechanics and heat transfer /
thermodynamics that is necessary to calculate how to properly cool a reactor.

Thermal-Hydraulics also dictates the temperature feedback to the neutron transport problem.

In essence, neutron transport, fluid flow, and heat transfer all become a multi-physics coupled system
in a reactor.

Dr. Gregory Greenman
Physicist

 

[aliaze1]

do you guys know any graduate programs that do nuclear reactor/system design research?

 

[Astronuc]

Undergraduate and graduate programs usually have a specific reactor or power plant design course in which one applies what one has learned in the various preparatory courses in neutronics, thermal-hydraulics, and perhaps mechanics (of materials). Some design courses may include mechanics of pipes and pressure vessels.

Taking a Reactor System Safety Analysis before one takes neutronics or basic reactor theory may be problematic if the safety analysis course uses theory that one would learn in the reactor theory course.

 

[gmax137]

Having a Masters or PhD will probably make you more 'interesting' to the hiring managers at a company like W but it certainly isn't necessary. The best thing you could do is try to get into their summer intern program. Too late for this summer, but if you are going to grad school next year, I suggest you seriously look into it for next summer. You get to see what really goes on day to day in the work environment, and they get to see how you work. Believe me, this is the 'inside track' to getting hired. Book knowledge is of course important (esp in this field) but work ethic, and overall 'brightness' is more so. If you have the self motivation, school is just the start, you will learn much more on the job and you can continue learning throughout your career.

 

[Astronuc]

hmm I am not 100% sure yet. someone had to come up with the CANDU idea, or the PWR idea, or any of the various submarine/boat ideas. that type of position...if i were to be more specific i guess i'd do reactor safety/system safety stuff

Gas (Magnox, AGR), CANDU, PWRs/VVERs and BWRs, which are the bases of current commercial power plants evolved during the 1950s-1970s. There were other, more exotic desings that didn't make it. By the 1970's designs were more or less fixed. The current advanced LWRs use the current fuel designs, although there are some proposed innovations in fuel element design.

The next generation of nuclear power plant (Gen IV) require new materials and slightly modified fuel designs, but core configurations are more or less unchanged.

 

[aliaze1]

Undergraduate and graduate programs usually have a specific reactor or power plant design course in which one applies what one has learned in the various preparatory courses in neutronics, thermal-hydraulics, and perhaps mechanics (of materials). Some design courses may include mechanics of pipes and pressure vessels.

Makes sense.

Taking a Reactor System Safety Analysis before one takes neutronics or basic reactor theory may be problematic if the safety analysis course uses theory that one would learn in the reactor theory course.

I spoke to the professor who teaches the course, he showed me the course website which had the following information:

topics covered:

1. Overview
2. Natural Disasters and Man Made Accidents
3. Safety Definitions and Terminology
4. Accidents Occurrence
5. Risk Quantification
6. Incidence and Likelihood Risk and Safety Indices
7. The Risk Assessment Methodology
8. Risk and Safety Ethics
9. The Source Term
10. Decay Heat Generation in Fission Reactors
11. Cost Effectiveness Analysis
12. Boolean Algebra
13. De Morgan Fuzzy Algebra
14. Probabilistic and Possibilistic Fault Tree Analysis
15. Random Numbers Generation
16. Direct Simulation or Analog Monte Carlo
17. Sampling Methods
18. Sampling Special Distributions
19. Event Tree Analysis
20. Fluid Mechanics Equations
21. Computational Fluid Dynamics
22. Autonomous Battery Reactors
23. Fourth Generation Reactor Concepts
24. Inherently Safe Reactor Designs
25. The Three Mile Island Accident
26. Chernobyl Accident
27. Global Climatic Change and Energy Use

 

[Astronuc]

That seems like a PRA course rather than the kind of reactor/plant safety analysis I took. In that case, we considered the core kinetics and plant dynamics in detail, with less emphasis on PRA.

 

[aliaze1]

my friends in nuc.E said that quantum mechanics do come into play, but not as much as one would think...they do more with neutron diffusion/transport/stuff with neutrons in general. based on your experience, should i take quantum mechanics? i do believe it is a prereq for some upper level quantum mechanics courses, as well as the reactor theory courses, but aside from that is it useful for a nuc.E?

also, any suggestions on EE classes?

thanks

 

[Morbius]

my friends in nuc.E said that quantum mechanics do come into play, but not as much as one would think...they do more with neutron diffusion/transport/stuff with neutrons in general. based on your experience, should i take quantum mechanics?

aliaze1,

YOU BET! One of the most important feedback mechanisms that is important for reactor safety is
Doppler Broadening of Absorption Resonances.

Essentially what happens is when a reactor gets hot; Doppler Broadening serves as a negative feedback
and decreases reactivity. If you know about negative feedback from your EE or systems engineering
courses; then you know that negative feedback keeps something stable.

Neutron resonances are high thin "peaks" in the neutron cross-section [ reaction probability ] as a
function of energy as seen in the following graph:

http://www.nndc.bnl.gov/sigma/getX4.jsp?evalid=4480&mf=3&mt=102

See all those peaks in the region between 1 ev and 1000 ev? Those are the ( resolved ) resonances.

The reason the neutron (n, gamma) cross-section looks like that and has those resonances are
all due to quantum mechanical effects.

So if you are going to understand how this whole process works - and how it relates to the stability
and safety of a reactor - you are going to have to know quantum mechanics.

Dr. Gregory Greenman
Physicist

 

[Astronuc]

my friends in nuc.E said that quantum mechanics do come into play, but not as much as one would think...they do more with neutron diffusion/transport/stuff with neutrons in general. based on your experience, should i take quantum mechanics? i do believe it is a prereq for some upper level quantum mechanics courses, as well as the reactor theory courses, but aside from that is it useful for a nuc.E?

It's not possible to make a useful comment without a syllabus. If it's an introductory QM course, it might cover mostly at the atomic level, rather than nuclear. The Nuc E program I took had a nuclear physics course that covered relativity and basic QM during the sophomore year, but I'd already had exposure to QM in a physics program.

Personally, I believe nuclear engineers should have as much physics as possible.

We took several EE courses including ciruit analysis, electric machinery, and control theory (as an option). We have at least one Nuc E who double-majored as an EE.

 

[aliaze1]

It's not possible to make a useful comment without a syllabus. If it's an introductory QM course, it might cover mostly at the atomic level, rather than nuclear. The Nuc E program I took had a nuclear physics course that covered relativity and basic QM during the sophomore year, but I'd already had exposure to QM in a physics program.

It's a 400-level QM course, but it is taught by a Nuclear Engineering professor, not a Physics professor. My friends who have taken it said that it is your standard QM course, particle in a box, compton effect, shrodinger equation, double square well, etc.

This looks like a general outline:

Principle Topics Covered

Classical Theory of Charged Particle Cross Sections

>Rutherford Scattering of Alpha Particles and the Nuclear Atom

>Scattering of Charges Particles by Atomic Electrons and Stopping Power

>Limitations of Classical Theory and the Need for Quantum Theory

Quantum Mechanical Principles and Methods

>Postulatory Basis of Quantum Mechanics

>Probability Current Density

>Hermitian Operators, Eigenfunctions and Eigenvalues

>Time-Dependent, Nondegenerate Perturbation Theory

>Dalgarno Method for Second Order Perturbation Theory

>Time-Dependent Perturbation Theory and Transition Probabilities

>Degenerate Perturbation Theory

>WKB Approximation

>Fox-Goodwin Numerical Method for Second Order Ordinary Differential Equation

Elementary Exact and Numerical Solutions of the Schroedinger Equation

>Bound States

>Tunnel Effect

>Harmonic Oscillator

>Atomic Structure

>Electrons in Periodic Lattice

>Application of the Fox-Goodwin Method to the Radial Schroedinger Equation

Quantum Analysis of Cross Sections

>Born Approximation

>Distorted Wave Born Approximation

>Method of Partial Waves

>Golden Rule

Photon Interactions with Atomic Electrons and Nuclei

>Semiclassical Theory of Radiation

>Compton Scattering and Absorption

>Photoelectric Effect

>Pair Production

>Attenuation coefficients

Radioactive-series Decay

>Radioactive Families (4n, 4n + 1, 4n + 2, 4n +3)

>Differential Equations for Growth and Decay

>Integral Formulation for Growth and Decay

>Production of Short-lived Isotopies



If I may ask, where did you go to school for NE?

 

[Astronuc]

That looks like an excellent course to take! It seems to be a blend of radiation interaction with matter and QM, which is a good combination.

Most of the time, one does not need QM for reactor design, particularly in the area of cross-sections. All that has been done. There are proprietary packages from the vendors and few independents that collapse the multi-energy spectrum cross-section libraries into more manageable energy groups. Codes like CASMO do this. The core simulators collapse the energy groups further.

 

[gmax137]

I agree with astronuc - you will not spend your days at Westinghouse solving schrodinger's eq. But Morbius is also right, up above where he says you need to know why these cross sections behave as they do. Well, you don't *need* to know in order to run the codes. But if you're going to be one of the guys who understands what the codes are doing, you do need to know. Besides, that course description sounds like fun.

 

[Astronuc]

Some of the abstracts will give insight into what the NSSS/reactor suppliers are considering.

http://www.inspi.ufl.edu/topfuel2009/program/ [Broken]
See - Advances in Water Reactor Fuel Technology and Testing

This paper is a good summary - Codes & Methods Supporting AREVA Fuel Solutions for the Future - Development Strategy
http://www.inspi.ufl.edu/topfuel2009/program/abstracts/2191.pdf [Broken]

Westinghouse has a New Reactor Core Engineering (NRCE) group, and AREVA has their EPR Project group.

GE is working with Hitachi with their ABWR and ESBWR.

 

[Morbius]

I agree with astronuc - you will not spend your days at Westinghouse solving schrodinger's eq. But Morbius is also right, up above where he says you need to know why these cross sections behave as they do. Well, you don't *need* to know in order to run the codes. But if you're going to be one of the guys who understands what the codes are doing, you do need to know. Besides, that course description sounds like fun.

gmax,

Your statement above reminds me of a very apt maxim:

They guy who knows HOW to do something will always have a job...
Working for the guy who knows WHY!

To aliaz1:
Do you just want to be able to run codes?
Or do you want to know what you are doing and why you are doing it?

Dr. Gregory Greenman

 

[gmax137]

The guy who knows HOW to do something will always have a job...
Working for the guy who knows WHY!

That's a good one!

I have been thinking about this some more since I posted above - if all you want to do is run the codes, you don't really need to go to grad school. You can learn on the job *how* to run the codes. But when problems come up, or the customer starts asking "why..." then the code runners will probably fail open...

 

[Astronuc]

Given the maturity of cross-section libraries, ENDF/B-6 and now 7, QM is not something that one would likely employ in reactor design. In the current Gen 3/3+ designs, the LWR fuel designs are already developed. There are separate fuel development programs running in parallel.

Some of the core technology include vanadium incore detectors.

Reactivity control is an important area for development.

See - http://world-nuclear.org/info/inf08.html


Back in the 90's, one vendor was having trouble with meeting target eigenvalues with a particular fuel design using a particular burnable absorber in certain cores. That is an area where cross-section data and neutron flux energy spectrum knowledge is critical.

Two group diffusion theory is still the bases of core simulators. ANL is looking a advanced transport theory in the NNR program, and perhaps in time, the vendors will consider the technology. One vendor shelved their program back in the late 90's since it was too expensive.

 

[Morbius]

Given the maturity of cross-section libraries, ENDF/B-6 and now 7, QM is not something that one would likely employ in reactor design.

Astronuc,

I used to think that cross-section libraries were fully mature - but we keep our cross-section people
fairly busy updating our nuclear data.

For example, I recently did some calculations where aluminum was used as a "filter" to adjust the
energy of a beam of fast neutrons down to the epi-thermal levels that were desired. I ran the
calculation with both an ENDF/B-6 and an ENDF/B-7 cross section sets and got very different
results .

It turns out that there are resolved elastic scattering resonances that are in the ENDF/B-6 libraries
that are not even in the ENDF/B-7 libraries. You can even use Brookhaven's cross-section
plotting software to compare the data and see the differences - look in the neighborhood of 10 keV.
I still have our cross-section people attempting to resolve this.

Dr. Gregory Greenman

 

[Astronuc]

Astronuc,

I used to think that cross-section libraries were fully mature - but we keep our cross-section people
fairly busy updating our nuclear data.

For example, I recently did some calculations where aluminum was used as a "filter" to adjust the energy of a beam of fast neutrons down to the epi-thermal levels that were desired. I ran the calculation with both an ENDF/B-6 and an ENDF/B-7 cross section sets and got very different results .

It turns out that there are resolved elastic scattering resonances that are in the ENDF/B-6 libraries that are not even in the ENDF/B-7 libraries. You can even use Brookhaven's cross-section plotting software to compare the data and see the differences - look in the neighborhood of 10 keV. I still have our cross-section people attempting to resolve this.

Dr. Gregory Greenman

Interesting. I seem to remember similar issues with previous versions from 4 -> 5 -> 6.

I think the development of ENDF is left to the DOE, and various national research organizations. The commerical institutions are cost-conscious and wherever possible, they like the government to pay for the R&D.