Designing a Core Lattice: Thermal-Hydraulics & Neutronics Considerations

[candice_84]

When designing a core lattice, which one of the thermal-hydraulics and neutronics consideration are important?

 

[Astronuc]

Neutronics, or rather, power distribution goes hand in hand with thermal hydraulics.

Basically one has to do the nuclear design subject to the constraint imposed by the thermal hydraulics (thermal boundary conditions).

There's an interplay also with the moderation, moderator temperature coefficient, in addition to the critical heat flux (CHF) and the margin to CHF.

One also has to be mindful of 10 CFR 50, App. A and the General Design Criteria, particularly GDC 10, 27 and 35, which are frequently cited in NUREG-0800, Standard Review Plan, particularly Section 4.2, Fuel System Design.

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appa.html

Reactor
http://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/sr0800/ch4/

4.2 Fuel System Design
4.3 Nuclear Design
4.4 Thermal and Hydraulic Design

 

[candice_84]

How about moderation-to-fuel ratio?

 

[Astronuc]

How about moderation-to-fuel ratio?

That's one parameter.

One has to do a balance between fuel pellet diameter/cladding diameter/ and rod pitch. One can open up the lattice which is cooler, but there is more leakage of neutrons.

Or for a given pitch, one can make a smaller fuel rod diameter, but that drives of the heat flux for a given linear power.

Of course, my example applies to a tubular design.

One can make a core a bit more homegeous with spherical fuel particles which pretty much necessitates a gas like He, Ne, CO2 for heat transfer. Reactivity management can be more of challenge though.

Then again, one could used a liquid fuel, e.g., molten salt, but that requires special considerations of the primary circuit and heat exchangers.

 

[candice_84]

Is it true,If we increase the fuel, the fuel utilization increases? therefore moderation-to-fuel ratio decreases and the reactor becomes under moderated?

 

[Astronuc]

Is it true,If we increase the fuel, the fuel utilization increases? therefore moderation-to-fuel ratio decreases and the reactor becomes under moderated?

Yes, and that would harden the spectrum. In BWRs, one can reduce flow and increase voiding in an assembly and harden the spectrum. The harder spectrum produces more Pu-239 from conversion of U-238, which can increase fuel utilization. That operation is called spectral shift.

So-called standard 17x17 PWR fuel has a UO2 pellet diameter of 0.3225 inch (8.19 mm), and cladding OD of 0.374 inch (9.5 mm). The lattice pitch is 0.496 inch (12.6 mm).

In the 1980s, Westinghouse introduced an optimized 17x17 which used a slightly thinner fuel rod which used a pellet with an OD of 0.3088 inch (7.84 mm) and cladding OD of 0.360 inch (9.14 mm) in the same lattice. The same fuel rod dimension is used VVER-1000 fuel assembly, but the lattice is triangular/hexagonal.

With the fatter fuel rod, one can load more fuel into the core, which can be used to reduce enrichment (and SWU costs), or go for slightly longer cycle length/exposure.

 

[candice_84]

Is there any advantage for the Russian design to be hexagonal over PWR lattice, or they just wanted to make it look different?

 

[Astronuc]

With respect to advantage, one has to look at the fuel moderator ratio, and critical heat flux.

The pitch of the VVER-1000 fuel rod is 0.5 inch (12.7 mm), but that is on a triangular geometry rather than a square lattice geometry. One should also look at the hydraulic diameter, as well as the fast to thermal flux ratio, or relative proportions in different neutron energy groups.

I understand that the VVER lattice is well suited for thorium based plants.

 

[candice_84]

Where do they fit a flux detector in either pwr or vver? Is it inside the fuel rod? also does it conflict with absorber rods? Also when you mentioned about spectrum what spectrum exactly you are pointing on?

 

[Astronuc]

Where do they fit a flux detector in either pwr or vver? Is it inside the fuel rod? also does it conflict with absorber rods?

One can do both in-core and ex-core detectors. The problem with ex-core detectors is that the neutron flux is very low at the core periphery.

In-core detectors are usually placed in an instrument tube, and the incore instrumentation includes thermocouples for core inlet or core outlet. The instrument tube is like a guide tube, but it is usually centrally located in an assembly of odd-numbered rows (15 x15 or 17x17) or slightly off-set from center in 14x14 or 16x16 assemblies, except for CE plants which have 4/5 guide tubes displacing 4 fuel rods.

AREVA's EPR is designed to use a guide tube instead of an instrument tube, and the EPR fuel assembly has 265 fuel rods instead of the traditional 264 fuel rods of a 17x17 lattice. W's AP-1000 and Mitsubishi's US-APWR use the traditional centrally located instrument tube. The instrumentation is placed in an unrodded location, i.e., no control rods in the assembly. AREVA also uses a pneumatic activation analysis system, which is based on Siemens technolgoy, in addition to typical SPN detector strings.

Some plant designs had incore probes inserted from below the core, but the modern plants favor insertion from the top.

 

[Astronuc]

Where do they fit a flux detector in either pwr or vver? Is it inside the fuel rod? also does it conflict with absorber rods? Also when you mentioned about spectrum what spectrum exactly you are pointing on?

Neutron energy spectrum.

 

[candice_84]

Can we use the same 17x17 fuel assembly in fast reactors? also what materials are suited as an absorber for fast reactor, because I was checking the cadmium's cross section and noticed it doesn't absorb at high energy level.

 

[Astronuc]

Fast reactors generally use hexagonal lattices.

The reference structural material for fast reactors is SS316 - both core and fuel structures. Other steels have been developed for improved creep resistance and reduced swelling.

One can find information here on the Clinch River Breeder Reactor
http://www.osti.gov/bridge/servlets/purl/4281663-9ySuUJ/4281663.pdf

Otherwise there is EBR-I, II at INL (INEL).

One can search on EBR-II and FFTF driver fuel.

Unfortunately, when FFTF was mothballed and trashed, a lot of fuel material was dumped and a lot of information was tossed. :(

A good reference is the Proceedings from an ASTM conference: Effects of radiation on materials / Radiation-induced changes in microstructure.
Example: http://books.google.com/books?id=k5j2P1xKKqEC&pg=PA146&lpg=PA146

 

[candice_84]

Fast reactors generally use hexagonal lattices.[/url]

Since there is no moderator involved in fast reactors, I assume hexagonal lattices perform better thermal-hydraulics than rectangular 17x17 in general. Is this a right assumption? Can you provide me a reference that shows how to mathematically do thermal-hydraulics analysis?

 

[Astronuc]

Since there is no moderator involved in fast reactors, I assume hexagonal lattices perform better thermal-hydraulics than rectangular 17x17 in general. Is this a right assumption?

I believe that the tringular/hexagonal lattice is better from a neutronic standpoint. The neutron leakage is more before the mean free path of a fast neutron is much greater in a fast reactor than in a thermal reactor. In the latter, fast neutrons slow down relatively quickly.

The thermalhydraulics in a fast reactor is a bit different since it is liquid metal, Na or NaK, or if one wants to get really exotic, Li. One has to be concerned with the positive void coefficient.

I believe that the control elements were B₄C based. It's been a while since if reviewed the technology.

I'll see if I can find a good reference on fast reactor T/H. Probably Waltar and Reynolds is the best text on fast reactors.

 

[candice_84]

In the pdf file for LMFR it says that LMFR don't need emergency cooling systems but I don't understand it.

 

[Astronuc]

In the pdf file for LMFR it says that LMFR don't need emergency cooling systems but I don't understand it.

Probaby there is a decay heat removal system, but not an ECCS like that of an LWR. The liquid metal coolant has excellent thermal conductivity, and I suspect that in shutdown, there is an effective heat removal system and one does not have to be concerned of the core overheating or the coolant boiling as the case with an LWR. The problem for LWRs is that they have to operate under high pressure to maintain a liquid phase. Na systems operate at a few atmospheres. IIRC, a Na loop at 0.20 or 0.5 MPa (about 2 to 5 atm) provides much the same heat transfer coefficient as pressurized water at 2250 psia (15.5 MPa), 290 C (563 K).

 

[Astronuc]

This is a reasonable good reference for LWR fuel design data - http://www.neimagazine.com/journals/Power/NEI/September_2004/attachments/NEISept04p26-35.pdf

 

[candice_84]

How important is it, to get an accurate burnup calculation and couple it with thermal-hyraulics?

 

[Astronuc]

How important is it, to get an accurate burnup calculation and couple it with thermal-hyraulics?

It depends on the calculation.

For economics, it's important with respect to determining utilization.

For fuel performance, it's important to get local (nodal) burnup as well as radial distribution of burnup for accurately predicting PCI and transient behavior. Many of the materials properties, e.g., UO2 thermal conductivity and swelling are strongly dependent on local burnup, while cladding mechanical properties are fluence and temperature dependent, and cladding oxidation is time and temperature, and to some extent water chemistry, dependent.

There are certain regulatory and licensing requirements that are dependent on burnup.

It is primarily fuel performance and certain operational/safety limits (and margins) for which accurate modeling of thermal-hydraulics is important. Properties dependent on the coolant local coolant conditions are important.

 

[candice_84]

Is it bad for the reactor to have high burnup?

 

[Astronuc]

There are burnup limits on commercial fuel due to licensing constraints on the fuel behavior related to postulated accidents and activity release.

 

[candice_84]

What are the advantages and disadvantages of upward and Downward flow in PWR?

 

[Astronuc]

What are the advantages and disadvantages of upward and Downward flow in PWR?

Is the question referring to core flow or upper head and by-pass flow?

 

[candice_84]

I mean core flow.

 

[Astronuc]

I mean core flow.

OK. I thought so, but there is also a use of the terminology of downflow and upflow for the bypass region which relates to baffle-jetting in some PWRs. Usually downflow plants have been converted to upflow in the bypass.

OK - what can one say about flow of a liquid involved in heat transfer in terms of differential pressure (pressure drop), bouyancy, and coolant temperature?

 

[candice_84]

Well, I'd like to know about all of them: Pressure Drop, Buoyancy and Coolant Temperature. My own guess about pressure drop is that when the flow is downward inside the core, there would be less pressure drop therefore less pumping power which is an advantage of downward flow. But the actual flow in the pwr goes down in the downcomer and then flows upward which i don't think its a good design but there has to be a trade off that I am not aware of it.

 

[Astronuc]

If one has a pipe and one measures the pressure at point A and down stream at point B, what can one say about the pressure at point A relative to point B? Assume that points A and B sit in the same horizontal plane with respect to the local gravitational field.

What would be the effect of point A being higher than point B, or point B being higher than point A?

 

[candice_84]

Pressure is higher at Point B, because P=\rhogh. Since H is higher in point B, pressure is more at point B.

 

[Astronuc]

Pressure is higher at Point B, because P=\rhogh. Since H is higher in point B, pressure is more at point B.

What if A and B are at the same elevation (h), or Δh=0, and water is pump (forced convection) from point A to B?

If one has a closed loop of piping, including a pump, and the pump is pushing the water around the loop, where is the maximum pressure?

 

[candice_84]

I think pressure at the horizontal line is the same.
I cannot picture the second question, therefore I don't know where the pressure is high in the tube.

 

[Astronuc]

I think pressure at the horizontal line is the same.
I cannot picture the second question, therefore I don't know where the pressure is high in the tube.

The primary cooling circuit or loop of a PWR is more or less a closed system, with the exeception of lines use to infuse or extract water for cooling pumps, boric acid, coolant sampling lines. The primary cooling system consists of the reactor core and pressure vessel (PV) internals, hot and cold legs, steam generators (headers and tubing), and cross over legs from the steam generator to reactor coolant pump on the cold leg. The pressurizer is attached to one of the hot legs and provides pressure to the system.

The highest pressure in the primary circuit is not the pressurizer, but at the outlet of the reactor coolant pump (RCP). The further one moves down stream from the RCP, the lower the pressure. The lowest pressure is at the inlet of the RCP. It's a bit like a voltage source in an electric circuit, the greatest electrical potential is at the + terminal, and the potential falls as one moves downstream around the circuit until returns to the - terminal (Kirchhoff's voltage law). The pressure drop across the RCP from inlet to outlet equals the pressure drop around the entire loop (continuity).

The pressure drop across the reactor core is about 25-26 psid. After the water enters the RPV through the nozzle with the cold leg, it then flows through the downcomer, through the lower plenum, up through the core support plate, through the core (flow channels between fuel rods), up through the upper core guide structure (where control rods are located) where it turns out to the hot nozzles.

Across the core, there is a pressure drop (pressure drops because of flow resistance). The temperature increases from the inlet (~280-292°C) to exit (~320-330°C). The reactor components in contact with the coolant are at those local temperatures. There may also be nucleate boiling in the hottest (combination of local coolant enthalpy/temperature and heat flux). The density of the coolant also changes.

The hot and cold nozzles are at about the same elevation, or the cold nozzles may be slightly lower than the hot nozzles. The coolant density is a function of temperature. The static head in the cold leg is of course greater than the static head in the reactor where there is a temperature gradient - and the hottest temperature is at the core exit.

So what are the advantages and disadvantages of downflow in the core vs the conventional upflow.

Think about component operating temperature, thermal hydraulics in the core, and safety issues.

 

[gmax137]

... But the actual flow in the pwr goes down in the downcomer and then flows upward which i don't think its a good design but there has to be a trade off that I am not aware of it.

Aside from any of the hydraulic effects Astronuc is discussing, think about what would happen if one of the main coolant pipes were to break. The vessel design with the downcomer allows the safety injection system to refill the vessel and flood the core. If you eliminate the downcomer, you would need to have the cold leg piping penetrate the vessel bottom - and a break in that cold leg piping would just drain away any safety injection flow. (And if you had flow 'up', it would be the hot leg piping that penetrates the lower head).

I think this is the main reason for the downcomer design (though there may be other reasons, like having the water provide shielding to the vessel material). Notice that the ECCS rules specify breaks in the main coolant loop piping - a break in the vessel itself is not a postulated design basis event.

 

[Astronuc]

I didn't specifically call out the ECCS, but that comes under safety considerations.

One might be interested in 3.9.5 Reactor Pressure Vessel Internals in AP1000 DCD Rev. 17
http://adamswebsearch2.nrc.gov/idmws/doccontent.dll?library=PU_ADAMS^PBNTAD01&ID=083251077

The entire DCD can be found
http://adamswebsearch2.nrc.gov/idmws/ViewDocByAccession.asp?AccessionNumber=ML083230868

Chapter 4 contains the information on the reactor.
Chapter 3 Design of Structures, Components, Equipment and Systems
Chapter 5 Reactor Coolant System and Connected Systems

Earlier version of DCD is found at - http://www.nrc.gov/reactors/new-reactors/design-cert/ap1000.html#dcd

 

[candice_84]

Can you guide me through calculating actual nuclear plants efficiency?

 

[gmax137]

Can you guide me through calculating actual nuclear plants efficiency?

OK. You start, by telling us what you mean by 'efficiency.'

 

[Astronuc]

Can you guide me through calculating actual nuclear plants efficiency?

One could start with Carnot efficiency.

Somewhere you'd want to compare the MWt (thermal energy produced) vs the useful electrical (or turbine mechanical energy to generator) available to sell.

 

[candice_84]

Ok, for example ap1000 produces 3030 MWth and 1000 MWe. I'd like to know how do we start from 3030 and produce 1000 MWe. I am aware of the fact that is Rankine Cycle and we reuse the steam from High pressure turbine to low pressure turbine. I am interested to know which formula to be used step by step.

 

[Astronuc]

Ok, for example ap1000 produces 3030 MWth and 1000 MWe. I'd like to know how do we start from 3030 and produce 1000 MWe. I am aware of the fact that is Rankine Cycle and we reuse the steam from High pressure turbine to low pressure turbine. I am interested to know which formula to be used step by step.

Well one could take 1000/3030 = ~0.33 or 33% efficiency.

One has to look at the thermal power generated, which input into the primary coolant. So the thermal power must equal the mass flow rate times the change in specific enthalpy.

The energy is transferred through the steam generator, so one has the know the primary circuit mass flow rate and the decrease in the specific enthalpy of the primary coolant in the steam generator, and the mass flow rate and change in specific enthalpy in the secondary circuit.

Then one has to look at the turbines HP, IP (if applicable) and LP turbines. Some fraction (x) of the specific enthalpy will be lost through the HP turbine, and some fraction (1-x) through the IP and LP turbines. One has to know the inlet and exit conditions of the steam, and then the thermal to mechanical conversion efficiency in the turbines.

\dot{W} (power) out of a turbine = η\dot{Q}, where η is the efficiency and \dot{Q} is the rate of change of enthalpy, or \dot{m}Δh.

One could also use the conditions of the condenser, since much of the thermal energy lost directly to the environment is passed through the condenser.

So recent modern turbines enable PWR plants to approach 37% efficiency.

Also one has to differentiate between gross and net efficiency. The NPP must use some of the electrical energy it produces in order to drive pumps and operate various electrical systems.

 

[gmax137]

If you just want to calculate a value for the efficiency, I think Astronuc has given you a good start. The power companies that operate the plants generally think in terms of 'heat rate' which has units of Btu per kw-hr. This is the ratio of the heat input to the electric output. It is really the same concept as efficiency, only it is the inverse - higher efficiency gives a lower heat rate. You can see how the two are related by noting that one kw-hour is equivalent to 3412 Btu; 33% thermo efficiency would be a heat rate of 10,340 Btu/kw-hr.

If you want to know how the particular value comes to be (as a result of the plant design), well that is a much more detailed subject that can't really be explained fully in this format, it is more like the subject of a mechanical engineering course. But essentially, the plant architect-engineer is usually responsible for designing the so-called 'secondary side' or 'balance of plant.' The design of the entire Rankine cycle is typically represented on a drawing called a 'heat balance diagram.' This shows the path of the steam from the reactor (BWR) or steam generator (PWR) through the turbines and condenser and back through the feedwater system. It includes all of the extraction steam lines and flow paths through the feedwater heaters, as well as the reheat steam between the high and low pressure turbines. The flow rate and thermodynamic state (pressure and enthalpy) at the exit of each component is determined based on the design characteristics of the components. The details of the calculations (for a FW heater, for example) would be based on parameters such as number of tubes, tube area, tube thickness & conductivity, etc., along with the flow rates on the inside & outside of the tubes. The designer's job is to juggle all of this in a way that gives the lowest heat rate, keeping in mind the capital cost to build the unit and the equipment operating and maintenance costs.

I suspect there are people on physicsforums that know a lot more about it than I do; but you might have better luck finding them in the General Engineering section than here in Nuclear land.

 

[candice_84]

I am actually interested in the balance of plant section :) I am studying thermodynamics and nothing is more interesting than learing how to do this calculation...

 

[gmax137]

I think you will have better luck getting posts from the people who really know about designing the rankine cycle if you ask this (as a new thread) in the General Engineering forum - those guys probably haven't tuned into this thread ("Reactor Design").

 

[candice_84]

Does small lattice pitch make -reactivity?

 

[Astronuc]

Does small lattice pitch make -reactivity?

Reactivity within a fuel lattice is a function of enrichment (i.e., concentration of fissile nuclides), the geometry and composition of structural material and coolant, concentration of burnable (depletable) neutron absorbers, and temperature. Also, in an LWR, pressure and temperature of the coolant determine its density which has an effect on moderation and neutron spectra.

There is an optimal pitch in which reactivity is maximized. In an LWR, a smaller than optimal pitch would produce undermoderation, while a larger pitch would produce overmoderation.

 

[candice_84]

I have heard that in some reactors by reducing the size of lattice pitch, one could reduce the positive reactivity.
And it actually make sense because the neutrons that are born in the fuel, will have less moderation (since the lattice is small) and I assume that a little bit of moderation from the small lattice, will keep them in the range of resonance absorption energy range. Therefore all the neutrons that are lowered only to the resonance energy range, after passing the lattice region they will reach the neighboring fuel which has U238, and since the neutrons are in resonance region and U238 have high resonance at that range it will absorb those neutrons, therefore we get negative reactivity as a result. Am I right about this?

 

[Astronuc]

LWRs are optimally designed to moderate fast fission neutrons from energies of about 1 to several MeV to thermal energies, ~ 0.024 MeV. If one reduces the pitch while keeping the rod diameter the same, then there is less moderation, and the spectrum becomes a little harder, i.e., the energy spectrum would shift up a little. Looking at the cross-section data,

U-235
total sigma - http://www.nndc.bnl.gov/sigma/getPlot.jsp?evalid=4551&mf=3&mt=1&nsub=10
total fission sigma - http://www.nndc.bnl.gov/sigma/getPlot.jsp?evalid=4551&mf=3&mt=18&nsub=10

U-238
total sigma - http://www.nndc.bnl.gov/sigma/getPlot.jsp?evalid=4554&mf=3&mt=1&nsub=10
total fission sigma - http://www.nndc.bnl.gov/sigma/getPlot.jsp?evalid=4554&mf=3&mt=18&nsub=10

One can see where the fission cross-section dominates and where absorption (non-fission) dominates. The primary effect for an undermoderated system is the fact that the fission cross-section has 1/v dependency, i.e., as the neutron energy increases the fission cross-section decreases, and so the probability of non-fission absorption increases.

The fission process is controlled by virtue of delayed neutrons, which are neutrons that are emitted from various fission products. Since some neutrons are delayed by several seconds, these enable an LWR system to change power in a controlled manner. Prompt criticality is to be avoided, and introduction of more than $1 of reactivity is also to be avoided.

 

[candice_84]

Delayed neutrons are very few in compare to prompt neutrons, but how could they be so effective in controlling the reactor. When neutrons fission, 2.47neutrons are released. most of those neutron are 95% prompt. By the time the delayed neutrons emission (few millisecond), the prompt neutrons would have many fission chain cycle. so how could that delayed neutron be so effective in terms of control?

 

[Astronuc]

Delayed neutrons are very few in compare to prompt neutrons, but how could they be so effective in controlling the reactor. When neutrons fission, 2.47neutrons are released. most of those neutron are 95% prompt. By the time the delayed neutrons emission (few millisecond), the prompt neutrons would have many fission chain cycle. so how could that delayed neutron be so effective in terms of control?

Delayed neutrons account for about 0.0065 or 0.65% of the neutrons in the given population. Basically the reactor goes critical with delayed neutrons. They last long enough to effectively increase the mean lifetime of neutrons which allows for controlled power changes.

That fraction (0.0065) of neutrons is also equivalent to $1 of reactivity. If one adds $1 of reactivity to the system, i.e., k ~ 1.0065+ or k_prompt > 1, the reactor becomes prompt critical and the power ascension can be dramatic - the power multiplies rapidly in fractions of a second - faster than the reactor can scram.

A reactivity insertion accident is a big deal (potentially significant exposure of plant personnel and increased risk to public, and damage or disruption of the core) and that's why there are strict controls on reactivity increases.

Hopefully when one takes a course in reactor kinetics, it will become apparent and very well appreciated!

 

[candice_84]

Do fast reactors lower the amount of waste? since U238 fissions. but we only convert a little portion of the U238 to energy.

 

[Astronuc]

Do fast reactors lower the amount of waste? since U238 fissions. but we only convert a little portion of the U238 to energy.

Fast reactors produce fission products as well. They can push fuel to higher burnups - maybe 50-100 GWd/tHM (HM = heavy metal) or 5-10% FIMA (fission(s) per initial metal atom), and even higher. If one considers spent fuel, then fast could produce more energy for the same amount of waste. On the other hand, fast reactor fuel is usually of higher enrichment, so there is generally more up front DU produced for a given mass of fast reactor fuel. The selling point of fast reactors is the recycling of Pu and transuranics - but that complicates the fuel cycle because the fuel then has to be manufactured and handled remotely - which makes it much more expensive than conventional LWR fuel.