ne.gif (2791 bytes)     NE571 Reactor Theory and Design
Fall semester 1998

Return to Course Outline

 

SCALE sequence SAS2H reading assignment

In the SCALE sequence SAS2H manual (SCALE section S2), read sections S2.1 and S2.2.  Also take a look at the input section S2.5, although we are going to use an EXTREME subset of the input (and I will walk you through the parts we need). 
 

Performing Assembly Calculations with SAS2H

As you saw in the above reading assignment, the SAS2H SCALE sequence allows us to calculate -- in a somewhat limited, but usable, way -- the behaviour of a reactor ASSEMBLY as a function of reactor exposure.  I said LIMITED because the actual transport calculations performed are "only" 1D cylindrical geometries, but I said USABLE because an amazing amount of the neutronic behaviour of a  reactor lattice can be picked up with this model, if we are clever.  Learning this cleverness is, of course, the trick. 

SAS2H spatial assembly model

The overall pattern goes like this.  We start with a reactor assembly that is made up of a regular array of various types (hopefully not too many different types) of pin-cells.  For our example, we will consider the 17x17 Sequoyah array consiting of: 
  • 264 fuel pins of fixed (beginning-of-life [BOL]) enrichment, dimensions, and pitch
  • 24 "extra" positions that are boron poison pins in the first cycle and water pins (+ control rods) thereafter -- with their associated geometry
  • 1 instrument tube in the center
  • a "superstructure" that applies to the assembly as a whole: the coolant water, support grids, assembly walls, coolant water between assemblies, etc.

(NOTE: the above layout includes divisions between the pin-cells that are NOT physical.  They were included just to illustrate the "surrounding coolant" associated with each pin) 

The first of the coupled 1D cylindrical discrete ordinates calculations is the pin-cell calculations for the DOMINANT fuel-bearing pin type in the assembly -- in this case, obviously, the 264 fuel pins-- along with its "share" of the surrounding coolant. (The square coolant region is "rounded" into an equal-area circle so that the pin-cell calculation is a proper 1D cylinder.)  An outer white boundary condition is applied, modeling an infinite lattice of this geometry. 

The result of this step is a set of "smeared" cross sections for the pin cell (which automatically becomes Material ID 500), which turns our assembly into this: 



The second of the coupled 1D calculations consists of: 

  • one of the secondary pins (i.e., one of the black ones) surrounded by
  • its "share" of the coolant (represented by the white region, "rounded off" into a cylindrical annulus) surrounded by
  • its "share" (=1/24th for our example) of the smeared primary pin-cell material (green) surrounded by
  • its "share" (=1/24th for our example) of the "superstructure" material (other than the coolant) -- the outside assembly walls and the water that separates assemblies, control rod blades between assemblies, etc.

Again, an external white boundary condition is applied, modeling an infinite lattice.  (An external buckling equivalent to a 2 meter height -- which allows for both axial and radial buckling.) 

The theory is that this cell -- with the original pin-cell behaviour reproduced by the smeared region -- will have neutronic behaviour similar to that of the assembly as a whole.  (SAS2H keeps up with the fact that the volume of this cell is less than the volume of the assembly and scales the powers, reaction rates, etc., accordingly.) 
 

Burn-up in SAS2H

The second real strength of the SAS2H sequence is the fact that it uses the state-of-the-art depletion code, ORIGEN,  to calculate changes in the assembly material contents as the reactor operates. In the subset of SAS2H input that we will be using, you have to supply: 
  • fuel pin-cell material and geometry (in the LATTICECELL form, like in the previous exercise)
  • the number of reactor cycles (= operation periods followed by shutdown times)
  • the number of burnup steps per cycle (for calculation purposes only)
  • several other variables describing the assembly: fuel length, print level, input level.
  • material and geometry for the second 1D model (either once or for each burnup step)
The output of SAS2H includes the reactivity history, the material contents versus reactor exposure, and (if you ask for it, which we will do later) few-group assembly cross sections for each burnup step. 

Okay.  Enough of this.  Let's put together a calculation or two. 

First reactor assembly calculation

For our first assembly calculation, let's just build an assembly out of the pin-cells we calculated in the previous example.  It will be interesting to see how long such an assembly would remain critical, plus it will give us an easy example to start with. 

We begin with the actual input from the previous exercise, with the first line changed to invoke SAS2H and include a special parameter that "turns off" the shipping cask calculation (that SAS2H was actually written for): 

=SAS2H     PARM='SKIPSHIPDATA
SEQUOYAH 2.6% FUEL PINS 
44GROUPNDF5                 LATTICECELL 
UO2            1  DEN=10.xx 1.0 8xx 92235 2.6 92238 97.4 END
ZIRCONIUM      2  1.0 6xx END
H2O            3  DEN=.7xxx 1.0 5xx END
B              4  DEN=.7xxx 1.0 5xx END
END COMP
SQUAREPITCH 1.xx .8xxxx 1  3  .9x 2  .8xxxx 0  END
END

These lines completely describe the primary pin cell.  Now we add a line of general information before the final "END" card: 

NPIN/ASSM=279 FUELNGTH=366 NCYCLES=1 NLIB/CYC=8 PRINTLEVEL=4 
          INPLEVEL=2 NUMZTOTAL=2 END

where 

  • NPIN/ASSM = number of fuel pins per assembly (For this exercise, we have filled the 17x17 with fuel pin cells
  • FUELNGTH = length of fuel pin in centimeters  (Our fuel rods are 12 feet long)
  • NCYCLES = number of fuel cycles
  • NLIB/CYC = number of burnup calculations (=cross section libraries produced) per cycle
  • PRINTLEVEL = print level.  We will use 4.
  • INPLEVEL = socalled SAS2H "input level".  We will be using 2 for a while.
  • NUMZTOTAL = number of "rings" in the 2nd model, which will follow
Another variable, MXREPEATS, is included and set to 0 if you want to input multiple versions of the second model (which you will want to do if the assembly geometry changes for different cycles).  In this example, we are NOT changing the assembly geometry with burnup, so we leave the variable out.  Note that these data can spill over onto more than one card. 

Next, we input the second model (once only since we did not put the MXREPEATS=0 in the previous set).  In this case, there is no "other pin cells", so the second model is just a homogeneous glob of the smeared material from the primary fuel cell: 

0 0.01 500 1.000 

(This is material 500 -- the smeared material from the first model with an outer radius of 1 cm. Of course, the outer radius doesn't matter, since it is homogeneous.  The strange inner void region is necessary in SCALE4.4, for reasons I don't understand.) 

POWER=17.5 BURN=1200 DOWN=15 END

Finally, we add the burnup data for each cycle.  This input sets the assembly power at 17.5 MW, burns for 1200 days, and then has a 15 day down-time.  (I just made up the 15 to have something there.) 
 

Complete the assignment:

Create the deck as we have detailed in this lesson, run it and see how long the cycle will remain critical.  (How hard could it be?  I did all the work for you.)  Be prepared that the calculation runs a long time (about 35 minutes on my 133 MHz laptop) and creates a large output file.  So, if class starts in 15 minutes, you are out of luck. ;) 
 

E-mail me the following values versus assembly burnup (in MWD/MTU): 

  • reactivity (k-effective)
  • grams of U-235 per assembly
  • grams of Pu-239 per assembly
  • average flux level

Hints: 

1.  The unit MWD/MTU is "megawatt-days per metric ton of uranium".  A metric ton is, of course, 1000 kg.  (The assembly should start out with about half a metric ton -- calculate the exact amount from your Time=0 day contents in the output.) 
2. You can find the k-effective values by searching on "lambda".  This will, however, give you BOTH the pin-cell k-effective AND the lattice k-effective.  I am only interested in the lattice k-effective. 
3. To get the grams of various isotopes, search on "gram atom", make sure that the following line gives the "BASIS" as a "SINGLE REACTOR ASSEMBLY", and convert to grams.  (A "gram atom" is the same thing as a mole, i.e., 1 gram atom of U-235 is about 235 grams.) 
4. The flux is given at the bottom of the "gram atom" table. 

mailbox.gif (2541 bytes)Subject: Assembly#1 


Return to Course Outline                                                                                               © 1998 by Ronald E. Pevey.  All rights reserved.