Coupling of MASH-MORSE Adjoint Leakages with Spaceand Time-Dependent Plume Radiation Sources

Transcription

1 ORNL/TM-2000/335 oupling of MASH-MORSE Adjoint Leakages with Spaeand Time-Dependent Plume Radiation Soures. O. Slater J. M. Barnes J. O. Johnson J. P. Renier R. T. Santoro

2 DOUMENT AVAILABILITY Reports produed after January 1, 1996, are generally available free via the U.S. Department of Energy (DOE) Information Bridge. Web site Reports produed before January 1, 1996, may be purhased by members of the publi from the following soure. National Tehnial Information Servie 5285 Port Royal Road Springfield, VA Telephone ( ) TDD Fax Web site Reports are available to DOE employees, DOE ontrators, Energy Tehnology Data Exhange (ETDE) representatives, and International Nulear Information System (INIS) representatives from the following soure. Offie of Sientifi and Tehnial Information P.O. Box 62 Oak Ridge, TN Telephone Fax Web site This report was prepared as an aount of work sponsored by an ageny of the United States Government. Neither the United States Government nor any ageny thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the auray, ompleteness, or usefulness of any information, apparatus, produt, or proess dislosed, or represents that its use would not infringe privately owned rights. Referene herein to any speifi ommerial produt, proess, or servie by trade name, trademark, manufaturer, or otherwise, does not neessarily onstitute or imply its endorsement, reommendation, or favoring by the United States Government or any ageny thereof. The views and opinions of authors expressed herein do not neessarily state or reflet those of the United States Government or any ageny thereof.

3 ORNL/TM-2000/335 omputational Physis and Engineering Division oupling of MASH-MORSE Adjoint Leakages with Spaeand Time-Dependent Plume Radiation Soures. O. Slater J. M. Barnes J. O. Johnson J.-P. Renier R. T. Santoro Date Published: April 2001 Prepared by the OAK RIDGE NATIONAL LABORATORY Oak Ridge, TN Managed by UT-BATTELLE, LL, for the

4 U.S. DEPARTMENT OF ENERGY under ontrat DE-A05-00OR22725

5

6 ONTENTS Page LIST OF TABLES...v LIST OF FIGURES... ix ABSTRAT... xi 1.0 INTRODUTION RIGOROUS METHOD OUPLING INDIRETLY WITH PLUME SOURES Desription Testing of the Method Test ase Soure Building alulations Results Test ase Soure Building alulations Results Sensitivity GENMASH vs TORT Free-Field Dose Rates omputation Time Test ase Soure Building alulations Results APPROXIMATE METHOD OUPLING DIRETLY WITH PLUME SOURES Desription Testing of the Method omparison of DR4 and GENMASH Free-Field Dose Rates ONLUSIONS...17 iv

7 5.0 REFERENES...19 APPENDIX A. INPUT INSTRUTIONS FOR VARIOUS ODES...65 A.1 Input Instrutions for DR A.2 Input Instrutions for GENTORT...72 A.3 Input Instrutions for PLATEQ...74 A.4 Input Instrutions for VIST3DP...76 A.5 Input Instrutions for QUAD3D...77 A.6 Input Instrutions for G2GAIR...78 APPENDIX B. DESRIPTIONS OF FILES USED OR REATED BY VARIOUS ODES. 81 B.1 DR4 Diretional Fluene File Format...83 B.2 GENMASH Soure File Format...84 B.3 GENTORT Output File Format...85 B.4 QUAD3D Output File Format...91 B.5 G2GAIR Output File Format...92 B.6 PLATEQ Output File Format...93 B.7 VIST3DP Salar Fluene File Format...94 B.8 VIST3DP Diretional Fluene File Format...96 v

8 LIST OF TABLES Table Page Atomi Densities m ) of Elements in Mixtures Used in the Test alulations Extent of Strutures in the Geometry for TORT/DR3SRF Test Problem omparison of Dose Rates for Forward-Forward (TORT-TORT) and Forward-Adjoint (TORT-MASH) ouplings Energy Boundaries for the Photon Group Strutures Used in the alulations ANSI Standard Photon Dose Rates (rem/h) at Detetor Loations Inside a Two-Story onrete Building ANSI Standard Free-Field Photon Dose Rates (rem/h) at Detetor Loations Inside a Two-Story onrete Building ANSI Standard Photon Dose-Rate Protetion Fators at Detetor Loations Inside a Two-Story onrete Building omparisons of TORT and GENMASH Free-Field Dose Rates at X= km, Y=55 m, and Z= m TORT/DR3SRF Free-Field Fluene and Dose Rate and Protetion Fators at a orner Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (oupled neutron and photon adjoint) TORT/DR3SRF Free-Field Fluene and Dose Rate and Protetion Fators at a enter Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (oupled neutron and photon adjoint) TORT/DR3SRF Free-Field Fluene and Dose Rate and Protetion Fators at a orner Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (Photon only adjoint) Free-Field Fluene and Dose Rate and Protetion Fators at a enter Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (Photon only adjoint) Desriptions of the 3-D Quadrature Sets on the File reated by QUAD3D Free-Field Data and Protetion Fators for a Detetor at Position 1 in a Two- Story Building. The DR4 alulation used a 22-angle 1-D quadrature and an S 3-D quadrature to bin the 3-D fluenes at eight fluene points 8 ~ surrounding the building for eah time step (alulation time 23 min. for two time steps Free-Field Data and Protetion Fators for a Detetor at Position 2 in a Two- Story Building. The DR4 alulation used a 22-angle 1-D quadrature and an S 3-D quadrature to bin the 3-D fluenes at eight fluene points surrounding 8 the building for eah time step (alulation time 1.2 min. For two time steps beause of time savings from the Table 1 alulations)...36 vi

9 Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a twoa a story onrete building (ibrt=2 ; nrad=43 ; DABL69 23-group photon struture; time=38.8 min) Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a twoa a story onrete building (ibrt=2 ; nrad=82 ; SALE 18-group photon struture; time=34.2 min) Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a twoa a story onrete building (ibrt=2 ; nrad=43 ; SALE 18-group photon struture; time=34.1 min) Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a twoa a story onrete building (ibrt=3 ; nrad=43 ; DABL69 23-group photon struture; time=132.5 min) Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a twoa a story onrete building (ibrt=3 ; nrad=82 ; SALE 18-group photon struture; time=116.2 min) Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a twoa a story onrete building (ibrt=3 ; nrad=43 ; SALE 18-group photon struture; time=116.0 min) omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for an Off-enter Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by a 23-group photon library; attenuation data given for 43 distanes) omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for a entered Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by a 23-group photon library; attenuation data given for 43 distanes) omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for an Off-enter Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 82 distanes) omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for a entered Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 82 distanes) omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for an Off-entered Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 43 distanes) omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for a entered Detetor in the Ten-Story NGI Building D (inludes roof soure vii

10 ontribution; air attenuation by an 18-group photon library; attenuation data given for 43 distanes) omparison of fluenes and dose rates alulated by speial odes using GENMASH soures, the DR4 soure-integration proedure, and ANISN airattenuated angular fluenes in two group strutures omparison of GENMASH dose rates with TORT, DR4, and Test Program dose rates...50 viii

11

12 LIST OF FIGURES Figure Page 1. Exterior view of the two-story onrete building used in DR3SRF and DR4 test ases Exterior view of the roofless two-story onrete building used in the DR3SRF and DR4 test ases X-Y planar slie through building to show X-Y loations of detetors within the building. The first- and seond-floor detetors have the same X and Y oordinates Surfae plot of the energy-integrated soure s ) at 1-m height one hour after the start of a PWR aident Surfae plot of the energy-integrated soure s ) at 1-m height three hours after the start of a PWR aident D airborne plume plus ground radiation soure distribution as a funtion of distane along the X-axis D ground radiation soure distribution as a funtion of distane along the X-axis Geometry for the Large, Ten-Story, Thin-Walled, onrete Building Group 1 fluene spatial distribution as a funtion of quadrature and point soure size X-Y plane illustration of the original mesh used to integrate plume and surfae soure ontributions to a point at the enter of the enter box (subdivisions of the first twelve annular regions are shown) X-Y plane illustration of the revised oarser mesh used to integrate plume and surfae soure ontributions to a point at the enter of the enter box (subdivisions of the first eight annular regions are shown) omparison of DR3SRF and DR4 free-field fluene spetra at detetor position 1 in a two-story building for two time steps omparison of DR3SRF and DR4 shielded fluene spetra at detetor position 1 in a two-story building for two time steps...63 x

13

14 ABSTRAT In the past, forward-adjoint oupling proedures in air-over-ground geometry have typially involved forward fluenes arising from a point soure a great distane from a target or vehile system. Various proessing odes were used to reate loalized forward fluene files that ould be used to ouple with the MASH-MORSE adjoint leakages. In reent years, radiation plumes that result from reator aidents or similar inidents have been modeled by others, and the soure spae and energy distributions as a funtion of time have been alulated. Additionally, with the point kernel method, they were able to alulate in relatively quik fashion free-field radiation doses for targets moving within the fluene field or for stationary targets within the field, the time dependene for the latter ase oming from the hanges in position, shape, soure strength, and spetra of the plume with time. The work desribed herein applies the plume soure to the MASH-MORSE oupling proedure. The plume soure replaes the point soure for generating the forward fluenes that are folded with MASH-MORSE adjoint leakages. Two types of soure alulations are desribed. The first is a "rigorous" alulation using the TORT ode and a spatially large air-over-ground geometry. For eah time step desired, diretional fluenes are alulated and are saved over a predetermined region that enompasses a struture within whih it is desired to alulate dose rates. Proessing odes then reate the surfae fluenes (whih may inlude ontributions from radiation soures that deposit on the roof or plateout) that will be oupled with the MASH-MORSE adjoint leakages. Unlike the point kernel alulations of the free-field dose rates, the TORT alulations in pratie inlude the effets of ground satter on dose rates and diretional fluenes, although the effets may be underestimated or overestimated beause of the use of neessarily oarse mesh and quadrature in order to redue omputational times. The seond soure alulation uses a point kernel like method to alulate diretional fluenes at points surrounding user-speified detetor and struture loations and folds those fluenes with MASH-MORSE adjoint leakages to alulate free-field and shielded fluenes and dose rates and the indiated protetion fators. Both soure alulation methods were tested for determining dose rates at speified loations within a two-story onrete building. For two early time steps, results from the seond method are 50 to 60% higher than those from the first method, but the protetion fators differ by only a few perent. This indiates that the free-field and shielded fluene spetra are similar and differ mainly in magnitude. Other omparisons were made, and from the omparisons there was a pereived underpredition by TORT. That underpredition an be attributed to the oarse mesh and quadrature used in the alulations. The seond method, while not rigorously orret, requires muh less omputer time for a reasonably good estimate of the shielded dose rates within a struture. xii

15

16 1.0 INTRODUTION 1-8 For a span of more than thirty-five years, muh effort has been expended toward the alulation of shielded fluenes, dose rates, or reation rates within so-alled targets suh as buildings, armored vehiles, and other strutures or objets due to point radiation soures situated in low-density media (usually air, air-over-ground, or void in the ase of Ref. 1) at large distanes from the objets. It has been found that eonomial solutions an be obtained by separating the problem into three parts. In the first part, the geometry is simplified by removal of the target from the geometry, and a soure alulation is performed with a two- or three-dimensional disrete ordinates ode starting from a first-ollision soure file. For the ase of a target in air-over-ground medium, only air and ground media are inluded in the geometry for this problem. For the point soure, this alulation yields spae-dependent free-field fluenes for eah group in the hosen energy range. Also inluded in this part of the alulation is the preparation of the equivalent soure file and other files that will be used in the third part of the problem. In the seond part of the problem, an adjoint Monte arlo alulation is performed in a loalized and generally omplex geometry. The adjoint soure onsists of a point soure at the detetor loation and eah energy group has a soure strength of 1.0. Any response funtion (adjoint soure spetra) may be used, but the unit soure in eah group allows for more uniform sampling in energy. For eah leakage event, both the soure and leakage energy group numbers are written along with other parameters to an event file. This enables the alulation of differential spetra along with integral quantities at the detetors. The third part of the problem (the oupling alulation) determines the orrespondenes between the alulations in the first and seond parts of the problem (i.e. the spatial offsets and orientations) and folds the forward fluenes and adjoint leakages to alulate the shielded detetor response. Then shielding protetion and redution fators are alulated 5 based on the free-field and shielded dose rates at the detetor. The initial oupling ode, DR, used diretional fluenes that varied only in one spatial dimension. Subsequent extensions of the ode added more detailed spatial representation of the diretional fluenes. These ode extensions aommodated the newer methods and retained the DR method through the use an input parameter idos that auses the appropriate parallel oding to be exeuted. Reently, studies of the airborne transport of radiation plumes and the free-field dose rates that result 9 therefrom have been onduted. A series of odes is used to model radiation dispersal in plumes resulting from nulear aidents of various origins (suh as reators, nulear proessing plants, and weapons). Soures (mainly photons) resulting from airborne and ground-deposited radioisotopes are ombined by 10 the GENMASH ode into energy groups, and soure arrays are reated as funtions of time, spae, and energy. These data an then be integrated in point kernel fashion to give free-field dose rates at seleted loations. Unfortunately, the energy spetrum at the detetor is not obtained when the point kernel method is used. This was identified as a problem when a desire was expressed for using the plume soure in plae of the point soures previously used in air-over-ground forward-adjoint oupling alulations. Two methods, one rigorous and one approximate, were developed for implementing the plume soure in the oupling sheme. 1

17 Setion 2.0 desribes the rigorous method developed initially and results for test problems, while Setion 3.0 desribes the less time-onsuming approximate method. Setion 4.0 presents the onlusions. In addition, appendies inlude input desriptions for the various odes and the file formats used or reated by the various odes. 2

18 2.0 RIGOROUS METHOD -- OUPLING INDIRETLY WITH PLUME SOURES 2.1 Desription The rigorous method of solving a problem with a plume soure required an alteration of the form of the forward fluene file that one inputs to the oupling ode. First, the general nature of the plume soure neessitates that a three-dimensional (3-D) disrete ordinates alulation be performed to obtain the 11 free-field fluenes. The TORT 3-D disrete ordinates radiation transport ode was used for these soure 12 alulations. The GENTORT ode was used to prepare the soure input for the TORT ode from the GENMASH distributions. The energy struture of the soure was hanged from that of GENMASH to that of TORT. Postproessing odes TORSET, PLATEQ, and VIST3DP were used to prepare diretional fluene files (with or without plateout on the top of the struture) and free-field fluene files needed for the oupling alulation. Beause of the large file that would result if fluenes were saved for all the TORT mesh points, the mesh for whih the diretional fluenes were output by TORT were onfined to a relatively small path region of the geometry. A seond alulation, a MASH-MORSE adjoint Monte arlo alulation that is the same for both the rigorous and approximate methods, was also performed The oupling alulation used the ode DR3SRF (altered from a previous folding ode, DR3 ) to fold 5,7,8 the forward fluenes with the MASH-MORSE adjoint leakages. Sine the TORT and MASH-MORSE energy group strutures were different, it was neessary to hange the soure group struture from the TORT struture to the MASH-MORSE struture prior to oupling in DR3SRF. 2.2 Testing of the Method As stated above, the oupling alulation was performed using the DR3SRF ode, whih inludes several folding options seleted by the input parameter idos. The DR3SRF folding method (input parameter idos=2) was tested using three sample problems. For eah problem, dose rates were determined for detetor loations within multistory onrete buildings. The ompositions of all the material mixtures appearing in the building geometries are given in Table Test ase 1 The method was first tested by omparing alulated shielded dose rates at ten loations within a two-story onrete building Soure A fititious plume soure was used, and the spatial extent of that soure is desribed in Table 2. The energy spetrum of the soure was given by s(e) e &1.1E 3

19 where the photon energy, E, has units of MeV. The funtion s(e) is easily integrated over the group energy bounds to give the soure in eah energy group: E 2 s(e)de (e &1.1E1 & e &1.1E2 ) m E 1 where E 1 and E 2 are respetively the lower and upper energy boundaries (MeV) of the energy group Building The dimensions of the two-story building extend from 0.0 to m in the X diretion, from 0.0 to m in the Y diretion, and from!50.0 to m in the Z diretion (Z=0.0 m is at ground level). The floors and roof onsist of m-thik reinfored onrete slabs, the seond floor and roof being supported by m-wide m-high reinfored onrete beams running along the X dimension. Eah floor is divided into four offies and a entral hallway that runs along the Y dimension. The walls are omposed of m-thik masonry bloks and are represented in the model by redued-density ordinary onrete. Eah offie has one window on eah outside wall bordering it and a door along the interior hallway. The windows and doors are modeled with the same thiknesses as the walls. The windows are represented by low-density silion dioxide and the doors are represented by a low-density steel material. Figure 1 shows a drawing of the building with the windows and doors pitured as void for easy viewing in the figure. Figure 2 shows the building with the roof removed, and Figure 3 shows a planar slie through the building and small retangles at the X-Y oordinates of the seleted detetor positions (five on eah floor). The detetor X-Y oordinates are the same on both floors. The detetor position labeled number 1 is in front of a window and was intended to be faing the soure. However, beause the soure moves along the negative X axis, the detetor is atually turned away from the soure. Likewise, detetor 2 is shielded by a wall and is away from the soure. Detetor 3 is in the hallway while detetors 4 and 5 are behind doors in adjaent offies and were intended to be the most heavily shielded loations away from the soure. The detetors ould have been oriented as intended through a 180E rotation of the TORT free-field geometry about the origin during the soure alulation alulations. A 15-photon-energy-group, S10P 3 TORT free-field alulation was performed 17 using the fititious soure, and ten 23-photon-group (DABL69 group struture) MASH-MORSE adjoint alulations were performed, one for eah detetor loation within the building. For eah MASH-MORSE alulation, 2,000,000 histories were followed. alulations of dose rates at detetor loations within the building were performed by two methods. Both methods depend on TORT free-field fluenes on exposed surfaes of the building. In the first method, a building alulation is performed using TORT and the free-field fluenes desribed above. This is a TORT-forward to TORT-forward (TORT-TORT) oupled alulation. The seondary TORT alulation used an S 8 quadrature and P 3 Legendre polynomial expansions of the sattering ross setions. The seond method ouples TORT-forward free-field fluenes to MASH-MORSE adjoint leakages (TORT-MASH); DR3SRF performs the oupling for the latter method. Seventy TORT-MASH oupling ases were performed (seven time steps and 10 detetors). Boundary fluenes from the TORT free-field alulation were onverted to the 23-group struture (idif=1) in order to avoid rerunning the MASH-MORSE ases with the 15-group struture. 4

20 It was later disovered that some of the materials in the building were inorretly plaed in the geometry model in both the TORT and MASH-MORSE alulations. The air and ground mixtures were orret. However, the floors, roof, and supports should have been rebar onrete instead of portland onrete; the walls should have been onrete blok instead of steel; and the door and window mixtures should have been those shown in Table 1 instead of rebar onrete and onrete bloks, respetively. So, while the problem solved was not the one intended, both TORT-TORT and TORT-MASH alulations were performed using the same inorret materials. Hene, the results should be omparable Results. omparisons were made of fluenes and dose rates alulated using a forward-tort to forward-tort oupling (TORT-TORT) and those obtained using a forward-tort to adjoint-mash-morse oupling (TORT-MASH). The soure (surfae fluenes) for the seond TORT alulation in the TORT-TORT sequene was prepared in the same manner as the soure was prepared for the TORT-MASH oupling exept for the file header reords and the number of spatial mesh points (fewer were needed in the TORT-MASH ase beause interpolation was used to alulate fluenes at all other spatial loations). oupling alulations were performed both with and without plateout soures on the roof of the building. The diretional fluenes due to plateout were alulated simply by dividing the soure strength by the osine between a given diretion and the surfae inward normal. Results are shown in Table 3. Exept for two positions on the first floor, the agreement between results from the forward-forward and the forward-adjoint oupling is generally good to exellent. For the two results with the greatest disagreements (position 3 for soures A and B), similar results were obtained for building roof soures with or without the inlusion of plateout soures. Even some results with relatively large frational standard deviations are in good agreement. In addition, the results show the effet from the roof plateout soure to be up to 10% on the first floor and up to 25% on the seond floor. Results from this problem provided assurane that the oupling proedure was orret and ould be applied to a more realisti plume soure Test ase 2 The next test of the oupling proedure involved a more realisti GENMASH plume soure Soure. The ode GENTORT alulated TORT volumetri soures for eah of the seven time steps haraterizing the GENMASH soure file. Input instrutions for GENTORT are given in Appendix A. The GENMASH soure for this test ase was one due to a modeled hypothetial reator aident lasting three hours, and the soure file inluded data on the radiation distribution for several times after the aident initiation (1.0, 2.0, 3.0, 4.4, 5.8, 7.2, and 10.0 h). The GENMASH soure geometry extended from!80 km to +1 km in X, from!6 km to +6 km in Y, and from 0 km to +0.2 km in Z. The aident oured at the origin (X=0.0 and Y=0.0) with time-dependent releases between 0 and 3 hours. A strong unidiretional wind moved the radioative materials downrange (i.e. along the negative X axis). The 18 GENMASH group struture onsisted of 18 photon groups with energies ranging from 10 kev to 14 MeV (see the third olumn of Table 4), and the soure was regrouped into another photon 18-group 5

21 19 struture (TORT group struture ) shown in the fourth olumn of Table 4. Sine there was no soure in the energy range above 5.0 MeV, the top three groups of the TORT group struture were eliminated to produe a 15-group struture having a maximum energy of 5.0 MeV. Note that there is no soure in the top five GENMASH groups whih also have energies above 5.0 MeV. The TORT free-field geometry extended from!2 km to +2 km in X, from!1 km to +1 km in Y, and from!50.0 m to +1 km in Z. The mesh onsisted of mesh ells. The origin of the TORT system was plaed at (!8.0 km, 0.0 km, 0.0 km) in the GENMASH geometry. The mesh-edge soure file option was seleted. Figures 4 and 5 show typial plume soure distributions, and Figures 6 and 7 show one-dimensional distributions (integrated over two spatial dimensions) of the total and ground soures along the X diretion. The airborne soure, whih is not shown separately in these figures, is high near the reator loation at t=1.0 h and shifts toward!80.0 km at later times and ompletely disappears between 5.8 h and 7.2 h after the aident. The ground soure peaks initially near the reator but is nearly uniform from!10.0 km to!80.0 km after 4.4 h Building. The building is the same as that desribed in Setion exept for the orretions that were made in the material speifiations in MASH-MORSE alulations. TORT 15-photon-group free-field alulations were performed for the onverted GENMASH soures at the seven time steps, and boundary fluenes were saved at the surfaes of the building to be used later in oupling alulations. The TORT alulations were run using an S 8 symmetri quadrature, a 15-group photon energy struture (E < 5.0 MeV), and a P 3 Legendre polynomial expansion of the sattering ross setions. Following the ompletion of the TORT alulations, the diretional fluene files were proessed and optionally enhaned by ontributions from roof plateout soures, and the adjoint leakages and forward fluenes were folded together to give dose rates at the ten detetor loations used in the previous test problem. Adjoint MASH-MORSE alulations were performed for ten detetor loations within the building after the material speifiations were orreted (see Setion ). These MASH-MORSE alulations used the same photon energy struture (15 groups) as that used in the TORT alulations. For eah MASH-MORSE alulation, 2,000,000 photon soure histories were followed. Seventy oupling ases were performed with DR3SRF (seven time steps and 10 detetors). alulation times were 725 to 800 min for the TORT alulations, 61 to 82 min for the MASH-MORSE alulations, and 0.25 to 0.5 min for the DR3SRF oupling alulations. The TORSET and VIST3DP alulations that proessed the output TORT boundary fluenes (to be used by DR3SRF) used a ombined total of about 4.5 seonds Results. Photon dose rates inside the building as a funtion of time and detetor position are shown in Table 5. The free-field dose rates (i.e. the dose rates that one alulates at the detetors in the absene of a building) are shown in Table 6, and the dose-rate protetion fators are shown in Table 7. The dose rates show about the same trend for all detetors, rising slightly for the first three hours of the aident, falling somewhat more than two orders of magnitude about one hour after the aident stopped, and falling off slowly with time beyond 4.4 h. For the first floor, protetion fators rise slightly during the first three hours of the aident, dip somewhat after shutdown, and slowly rise at later times. The 6

22 seond-floor results are similar exept the protetion fators rise after shutdown instead of dipping as they did for the first-floor detetors. There are no independent results against whih to ompare these results. However, results from the first test problems give onfidene that the results are reasonable Sensitivity. The sensitivity of the TORT alulations to the spatial extent of the soure was tested. The TORT alulations were performed over a fration of the GENMASH geometry, and the size needed for the TORT geometry was determined using one-dimensional alulations. In those alulations, an airborne soure volume was added to a spherial geometry until the fluene at the enter of the sphere 5 stopped inreasing. From those results, it was determined that 2 10 m (2 km) of air with soure was suffiient. Thus, the TORT geometry extended from!2 km to +2 km in X. For time step 1, an additional TORT alulation was run with the geometry extending from!3 km to +3 km in X. The maximum inrease in the free-field dose rates was %. Hene, the smaller TORT geometry was deemed to be of adequate size. A seond sensitivity study examined the effet of the perturbations made to the soure in the region oupied by the building. Free-field and shielded dose rates were examined for perturbed and unperturbed soure fields. If the perturbation of the soure auses minor hanges to the dose rates, then the free-field TORT alulations an be performed without onsideration of the presene of the building or any other perturbation. The soure perturbation was implemented by redistributing the soure in the spae oupied by the building mokup uniformly in the mesh ells surrounding the building and distributing the ground soure as a volume soure in the mesh ells in the plane above the roof. If the soure perturbation an be ignored, then the ground soure ould easily be added to the roof boundary fluene prior to folding with the MASH-MORSE leakages. The omparisons of the shielded dose rates showed dose rates from the unperturbed field to be 2.2 to 8.3% lower on the first floor and 1.1 to 4.2% lower on the seond floor. There is more soure entering the building mokup in the perturbed ase beause about half of the isotropi soure, redistributed into the intervals surrounding the building, is direted into the building. That inreases the fluene that is folded with the MASH-MORSE leakages. While shielded dose rates were all lower for the unperturbed field, the effets on the free-field fluenes were mixed. For the loation in the ground beneath the building mokup, the dose rate is about 10% higher for the unperturbed field. Dose rates in the ground away from the building mokup differ little. The dose rate at a point within the building mokup and slightly above the ground is about 9% higher for the unperturbed field. For loations just outside the building mokup, dose rates for the unperturbed field ranged from 2% lower to 7% higher, most likely due to the assumption of a uniform redistribution around the building rather than a distribution like that for the intervals well outside the mokup. Dose rates above the roof were 1.4% lower to 1.5% higher for the unperturbed field. Overall, it appears that the unperturbed free-field fluenes will be adequate for alulating dose rates within buildings. So, if enough diretional fluenes are retained (i.e. over a large enough spatial extent), dose rates within the building may be obtained for any positioning of the building that leaves it within the extent of the fluene field. 7

23 GENMASH vs TORT Free-Field Dose Rates. The GENMASH ode alulates soures as well as dose rates due to those soures, and TORT uses reformatted GENMASH soures to alulate free-field fluenes in an air-over-ground geometry. The dose-rate alulations performed by GENMASH 20 use dose fators from IRP Publiation 51 and "Geometri Progression Gamma-Ray Exposure Buildup 21 Fator oeffiients" from ANSI/ANS 6.4.3, while TORT initially used ANSI standard dose fators from 19 the SALE library. Dose rates were reomputed from TORT salar fluenes using dose fators onverted from the IRP data. Results were ompared to GENMASH values and are shown in the Table 8. For the single spatial loation, TORT results are up to 40% below the GENMASH results omputation Time. The TORT alulations referred to above ran on IBM RIS/ workstations and in general required long omputation times (from 700 to 800 minutes for the oarsest mesh and quadrature used to over 1600 minutes for the finest mesh and quadrature used). The total PU time for the seven TORT 15-photon-group alulations was more than 5400 minutes (3.75 days). About 670 minutes of PU time were used to perform the ten two-million history, 15-photon-group MASH-MORSE alulations, and about 36 minutes were used for the folding. Thus, the total time expended was more than 6100 minutes (4.24 days). Obviously, the TORT alulation times may be onsidered exessive for the desired appliations. Hene, a simpler, less time-onsuming, alternate method was sought. This method is desribed in Setion Test ase 3 A final test ase involved the alulation of protetion fators at two detetor positions within a ten-story onrete building Soure. The soure is desribed in Setion Building. The ten-story building, shown in Figure 8, is m long (X diretion), m wide (Y diretion), and m high (Z diretion). Floors and walls are m thik, and there are window openings in the walls. The model also inludes a ground region that is m long, m wide, and m thik alulations. Adjoint MASH-MORSE alulations were performed using both oupled neutron and photon and photon-only DABL69 ross-setion libraries. Detetors were loated in a hallway orner (X=!840.74m, Y=30.48m, Z=1454.7m) and at the enter of the building (X=0.0m, Y=0.0m, Z=1454.7m) for the two MASH-MORSE adjoints. Four of the TORT ases desribed in Setion were rerun (1.0, 3.0, 5.8, and 10.0 h) after minor hanges were made in the spatial mesh in order to aommodate the large ten-story building (Setion ) within the revised and extended diretional fluene path boundaries. As with test ase 2, diretional fluene files were prepared for oupling with the MASH-MORSE adjoints. Fluenes and dose rates inside the building were alulated both with and without a plateout soure on the roof. 8

24 Results. Results for the oupled adjoints are shown in Table 9 for a detetor loated in a hallway orner (X=!840.74m, Y=30.48m, Z=1454.7m) and in Table 10 for a detetor loated at the enter of the building (X=0.0m, Y=0.0m, Z=1454.7m). Results for the photon-only adjoints are shown in Table 11 for a detetor loated in a hallway orner and in Table 12 for a detetor loated at the enter of the building. For these alulations, the adjoint soure was nonzero only for photons with energies below 5.0 MeV, although adjoint leakage was sored for all remaining groups of the oupled or photon-only struture. Of more than 180,000 leaking partiles for the oupled ases, only about 18,000 (or about 10%) atually ontributed to the alulated shielded fluenes and doses. For the photon-only ases, about one-third of the leaking partiles ontributed to the fluenes and doses. The results in Tables 9 and 10 show less than 10% variation in the protetion fators among the four dose funtions for any time step. The highest alulated protetion fators are those for time t=3.0 h. The tables also show that the plateout soure on the roof ontributes little in the first two time steps and about 2.5% in the last two time steps. Unertainties in the spetra for the oupled ases were less than 10% for all exept the last three groups (E < 45 kev). The unertainty for group 67 was about 14% while it was about 51% for group 68, whih ontributes little to the shielded fluenes and doses. Group 69 had no ontribution. The results in Tables 11 and 12 eliit onlusions similar to those for Tables 9 and 10. However, unertainties are greater than 10% for the last four groups (E < 70 kev) and differenes in the protetion fators between the oupled and photon-only ases are greater for the entered detetor (12 to 20%) than for the off-enter detetor (up to 4%). Additionally, for the last two time steps, the plateout soure on the roof ontributes 2.2% to the fluene but less than 1.5% to the dose rate. 9

25

26 3.0 APPROXIMATE METHOD -- OUPLING DIRETLY WITH PLUME SOURES 3.1 Desription A method of oupling diretly with a plume soure (input parameter idos=3) was developed in an effort to redue the overall omputation time required to produe the diretional fluenes to be folded with the MASH-MORSE adjoint leakages. The method is enoded along with all the previous DR3SRF folding methods in a ode alled DR4 for whih the input is doumented in Appendix A and the file formats are doumented in Appendix B. Before the method ould be implemented, two additional data files were required. One file is a 3-D quadrature data file with data in a form usable by the ode without further proessing. The quadratures are used to bin the diretional fluenes omputed by this method. The 22 file is produed by the QUAD3D ode and is organized suh that the number of diretions and the quadrature data for a given set are seleted based on a four-harater name for the quadrature set. For eah quadrature set in the file, Table 13 shows the set name, the number of diretions, and a brief desription. The seond file is one that ontains as a funtion of distane, the 1-D angular fluenes in various groups resulting from a soure (of whatever magnitude) in eah soure group. The data are treated as if there an be transfers from any one group to any other group, even though in most ases there will be only downsatter. To reate the seond file, one needs to run igm (number of energy groups) ANISN 23 alulations and save the angular fluene files. For the file reated to use with the test ases, alulations were run for 18 groups, but only the last 15 were used to reate the required file. Also, the air layer thikness was limited to about 5 km due to numerial problems on the IBM RIS 6000 workstations [ANISN is a single-preision arithmeti (32-bit) omputer ode. A double-preision ANISN might be more appropriate]. Additionally, the point soure was normalized to 10 s for eah group. This high normalization allowed most of the ANISN angular fluenes to fall within the range of allowable real numbers on the workstations. The ANISN alulations have to be performed for eah air omposition desired. The group struture of the ANISN alulations is assumed to be different from the GENMASH and MASH-MORSE group strutures. One should use a broad-group struture for the ANISN alulations beause disk- and ore-storage requirements are redued and the problem exeution time is shortened. Test ANISN alulations were performed for two point soure spherial radii (60 m and m) and S and S quadratures to determine if these parameters had muh effet on the total fluenes at the ranges of interest. Figure 9 shows a omparison of the group-1 fluene as a funtion of range for four soure ombinations. Those results show that outside the radius of the largest sphere, the results for the small and large point soures are not signifiantly different. Also, the two quadratures produed similar results. Therefore, the 60-m-radius point soure and the S 64 quadrature used in the ANISN alulations are adequate parameters. The ANISN angular fluene files were proessed through 24 the G2GAIR ode to ombine the data into a single file to be used by DR4. The user speifies the spatial loations (in asending order) at whih the angular fluenes will be output (these loations are modified by subtrating from eah the parameter rref speified by the user). The user also speifies broad angular bins (whih may be the same as or a subset of the input ANISN quadrature) into whih the output angular fluenes will be binned. Input instrutions for the QUAD3D and G2GAIR odes are inluded in 11 20!1

27 Appendix A and desriptions of the strutures of the files output by the odes are in Appendix B. In addition to these two files, the mesh-edged plume soure file reated by GENMASH is required. A desription of that file is also inluded in Appendix B. The DR4 alulational proedure (input parameter idos=3) is somewhat like a point kernel alulation, but the spatial attenuation is taken are of by ANISN, and the resulting group-to-group angular fluenes replae the buildup fators. While the point kernel method subdivides the soure volumes and alulates ontributions from eah subell, this method generates a relative mesh around eah point at whih angular fluenes are to be alulated. Sample X-Y plan views of the inner regions of the soure integration mesh are shown in Figures 10 and 11. Figure 10 shows the original relatively refined mesh, and Figure 11 shows a oarser mesh. The integration mesh represented by Figure 10 has about six times the total number of ells found in that of Figure 11, and test ode alulations showed small differenes in the fluenes alulated using the two integration meshes. There appears to be little additional ontribution from soure regions more than 2.5 km away whenever these are not the dominant soure regions. Therefore, the mesh in any oordinate diretion was limited by the GENMASH soure extrema and the maximum allowed distane from the fluene loation where ontributions are made to the diretional fluene. The initial values a of the diretional fluenes in a box about the fluene loation are alulated as the GENMASH soure value at the enter of the box times the average fluene one alulates by putting a uniform unit soure in the box and alulating the trak lengths per unit volume. It was found that soures within the box ontribute only a small amount to the total fluene at a given loation. Then, for eah of the soure integration meshes that fall within the established riteria, (1) the interpolation fator for the group gn to group g angular fluene is determined from the soure-to-detetor distane and where that distane falls within the G2GAIR output mesh; (2) the 1-D quadrature angles orresponding to eah of the 3-D quadrature diretions are individually determined; (3) the group gn to group g angular fluenes are determined through interpolation; and (4) a term equal to the produt of the soure, the interpolated group-to-group angular fluene, and the soure point volume or surfae leakage fator (area times the osine with respet to the surfae normal) is added to the diretional fluene for the appropriate diretions and groups. Free-field fluenes are alulated by interpolating between integrals of the diretional fluenes. However, it was noted that the free-field fluenes alulated in this manner were lower than those obtained using salar group-to-group transfers. This may possibly have been due to a lak of quadrature diretions falling within the most straight-ahead bin of the ANISN alulation. The original method for whih the 1-D quadrature sphere was finely subdivided and ontributions were summed into the transformed 3-D quadrature diretional bins probably would have given better agreement, but it appeared to be too time-onsuming. A remedy for this situation was the normalization of the diretional fluene ontributions for eah soure integration mesh to the group salar fluene ontribution based on a salar group-to-group transfer. Diretion m fluenes for ground points, and optionally roof points, are enhaned by the GENMASH ground soure value divided by the absolute value of eta(m), the Z-diretion osine. These a 50-m 50-m 50-m box for a point above the ground (ideally, the point should be more than 25 m above the ground) and a 50-m 50-m 25-m box for a point on the ground. 12

28 fluene enhanements may ause the interpolated diretional fluenes at the side walls to be slightly higher than they should be, sine one is interpolating between the higher fluene values on the roof and the unhanged values at points below the roof. The oordinates of the points [smesh(3,kpts,ndet)] at whih diretional fluenes are to be alulated are determined based on several fators. First, an input parameter, ibrt, gives the number of points per spatial diretion on a box surrounding the MASH-MORSE geometry system (minimum of 2). The total 3 number of points (kpts) is ibrt ubed (i.e. ibrt ). Seond the xo, yo, zo, and to input arrays give referene loations of the MASH-MORSE system within the GENMASH system and the time at whih the fluenes are to be alulated (provision is made for interpolating the GENMASH soures to alulate soures for input times that lie within the range of values on the GENMASH soure file). The Z extrema of the box for spae-time/orientation jdet are zo(jdet)+zmin and zo(jdet)+zmax, where zo(jdet) is the input ground surfae offset between the GENMASH and MASH-MORSE systems and zmin and zmax are minimum and maximum Z values of the MASH-MORSE geometry (set zmin suh that zo(jdet)+zmin=0.0). If ibrt > 2, then the other Z loations are equally spaed between the Z extrema. The X and Y oordinates are omputed suh that the MASH-MORSE geometry an be rotated 360 degrees and all parts remain inside the box. This is done by alulating the length of the X-Y diagonal of b the MASH-MORSE geometry and subtrating that from or adding that to xo(jdet) for the X extrema or subtrating from or adding to yo(jdet) for the Y extrema. For ibrt > 2, other X and Y loations are determined as with Z. Fluenes are alulated for kpts points for eah spae-time/orientation jdet, and the fluenes for any kdet > jdet that has the same spae-time oordinates as jdet are assigned the same fluene values as those omputed for spae-time/orientation jdet. This prevents the realulation of these quantities for ases having the same spae-time oordinates but different orientations. In addition, the DR4-alulated fluenes may be saved on a permanent or temporary file to be used by another ase in whih the only differene is a hange in the detetor position in the MASH-MORSE geometry and a orresponding hange in the leakage file used (a different file for eah detetor). A desription of the DR4 diretional fluene file is inluded in Appendix B. 3.2 Testing of the Method Two test ases for the plume soure oupling proedure are ases run with the DR3SRF idos=2 option (oupling with TORT boundary diretional fluenes as desribed in Setion ). The first test involves the alulation of free-field and shielded fluenes and dose rates at ten detetor loations in a two-story onrete building with windows and doors for seven time steps as was reported in Setion Results for the same time steps and detetor positions were obtained by the DR4 idos=3 proedure using about 123 minutes of PU time beyond the 670 minutes required for the ten previously-performed MASH-MORSE alulations. About 80 minutes of the DR4 time were used to alulate results for the first MASH-MORSE detetor position for nine spae-time/orientation b (xmax&xmin) 2 % (ymax&ymin) 2 13

29 ombinations. The results from this proedure do not inlude the effets of ground satter. A omparison of results obtained using TORT fluenes from alulations with and without the ground showed the ground effet to be (1) a 60% inrease in the free-field fluene, (2) a 20 to 30% inrease in the free-field dose rates, (3) a 20% inrease in the fluene protetion fator, and (4) a 10% inrease in the dose protetion fators. The low-energy photon groups ontribute more to the free-field fluene than they do to the other three quantities, thus justifying the larger ground effet for the free-field fluene. The idos=3 proedure of DR4 is only valid for large dispersed soures having relatively smooth transitions from a given region to adjaent regions. Effets of large loalized spikes in the GENMASH soure distribution may be missed beause of the fixed soure-integration mesh. Results from DR3SRF and DR4 for two detetor positions and two time steps are ompared in Tables 14 and 15, and fluene spetra are ompared in Figures 12 and 13. The DR4 fluenes and dose rates are about 50% higher than those alulated by DR3SRF, but protetion fators are in good agreement. As mentioned earlier, the DR4 results were on the order of the DR3SRF results before angular fluene ontributions were normalized by the salar fluene ontributions of eah soure integration mesh. However, with the normalization implemented, the ANSI dose rate is on the order of that alulated by GENMASH. One expets agreement with the GENMASH result beause both odes ignore ground effets. On the other hand, DR3SRF results should be higher than results alulated by GENMASH or DR4 with idos=3, sine ground satter enhanes the fluene levels. While the ground-satter effet doesn't seem to be very large for the photon-only soure, one would expet muh larger differenes for the oupled neutron-photon ase due to the ontributions of the seondary photons produed in the ground near the target. However, radiation plumes generally have very little or no neutron emissions. As mentioned earlier, the DR4 alulation may use data files in three different group strutures: GENMASH, MASH-MORSE, and ANISN group strutures. DR4 alulations were performed to test the effets of the number of rings (nrad) in the plume soure integration proedure and of the group struture used in the ANISN alulations. Results in Tables are for detetor position 1 on the first floor of a two-story onrete building. Free-field fluenes and dose rates and protetion fators are presented for three time steps and six variations on the input data. For Tables 16-18, eight fluene points surround the building (ibrt=2) and ANISN air-attenuated angular fluenes are for either (1) 43 spatial points and the DABL69 23-photon-group struture or (2) 43 or 82 spatial points and the SALE 18-photon-group struture. For Tables 19-21, twenty-seven fluene points (ibrt=3) define a box that enompasses the building, and other data variations are the same as those used for Tables First, the results show that over the first three hours, the fluenes, dose rates, and protetion fators inrease with time. The inrease in the protetion fators may be due to either a soure-spetrum shift with time toward lower energies or a spatial repositioning of the soure suh that a smaller soure fration has a nearly diret view of the detetor through a window. The seond observation to note is that inreasing ibrt from 2 to 3 inreases results by a few perent. A third observation to note is that the DABL69 23-photon-group struture gives higher results than the SALE 18-photon-group struture (A more refined photo energy-group struture than the one available in SALE might be more appropriate). Finally, using the 18-photon-group air-attenuation data file with 43 spatial points results in higher fluenes and dose rates than 14

30 those obtained using 82 spatial points, but the protetion fators are virtually the same. The seond test problem involves the alulation of dose rates and protetion fators for the ten-story building (Figure 8) used in the third test of the rigorous method desribed in Setion and subsetions. Results are shown in Tables for entered and off-enter detetor positions for three variations in the air attenuation file parameters (group struture and the number of air attenuation spatial data points). For the large ten-story building, DR4 tended to give fators of 1.5 to 2.0 higher fluenes than TORT-DR3SRF for the early times, but the DR4 results were smaller at later time steps (0.6 to 0.9 times the TORT-DR3SRF results). DR4 protetion fators were fators of 1.2 to 1.4 times the TORT-DR3SRF results. The use of ANISN air-attenuated angular fluenes alulated with the DABL69 23-photon-group struture (Tables 22 and 23) resulted in slightly higher (< 10%) fluenes and protetion fators than those obtained using ANISN air-attenuated angular fluenes alulated with the SALE 18-photon-group struture (Tables 24-27). Table 28 ompared free-field photon fluenes and dose rates alulated with a speial ode using the SALE 18-group struture to those alulated with a speial ode using the DABL69 23-group struture. The odes used the same soure integration proedure as DR4 but alulated results only at the speified points rather than the eight or more loations for whih DR4 alulates fluenes before interpolating to get fluenes at the points of interest. The fluenes for the latter struture were 3% higher and the Straker-Morrison dose rates were about 7% higher. These results are in good agreement. Unlike the results for the first test ase above, the protetion fators differ by as muh as 40% from the DR3SRF results. Perhaps, an inreased number of adjoint histories would improve the results. 3.3 omparison of DR4 and GENMASH Free-Field Dose Rates With regard to the omparison with free-field dose rates omputed with GENMASH (Table 8, Setion ), alulations were performed with DR4 and a test ode to see if the soure integration proedure would give similar results to GENMASH. The free-field dose rates were on the order of those alulated by GENMASH. However, the detetors were not at the same loations for whih the GENMASH dose rates were alulated (i.e. X=! km, Y=55 m, and Z= m, and t=1.0 h and 3.0 h). A test program, using the same soure integration proedure as DR4, ould alulate results at the preise loation, but DR4 alulates fluenes at several loations surrounding the detetor loation and interpolates on those values to obtain the result at the detetor. While DR4 uses the same soure integration method as the test program, the alulated dose rates may be different beause the DR4 result is not alulated at the preise loation. Previously, TORT results were obtained by interpolating fluenes from the TORT salar fluene file and folding the fluenes with the dose response funtion. Test program results were obtained for the detetor at the above mentioned loation, and DR4 results were obtained for four variations on the input parameters. The results are ompared to the GENMASH results in Table 29. As noted in the table, the TORT results are well below the GENMASH results. The test program gave results that are in good agreement with the GENMASH results. The input for the third DR4 alulation was suh that the fluene points were more losely spaed in X and Y than were those in the 15

31 first two ases. The fourth DR4 ase was like the third exept that the Z spaing of the fluene points was made less. This alulation effetively moves the ground soure loser to the detetor and may exaggerate the ground soure ontribution. It is noted that the results for the first DR4 alulation are slightly higher than the TORT results, but they are still below the GENMASH results by 25%. For this ase whih used ibrt=2, the X and Y extrema are more than 200 m apart, allowing peak fluenes in the box to be missed. The seond DR4 ase was the same as the first exept that ibrt was hanged to 3. This brought some of the interior fluene points loser to the detetor loation resulting in smaller errors due to interpolation. The results from this ase are in good agreement with GENMASH, being about 2% low at 1.0 h and about 9% low at 3.0 h. The third DR4 ase used the parameters shown in the footnote of Table 18. The X and Y extrema were brought loser to the detetor oordinates. The agreement improved slightly from ase 2 with the result being about 3% high at 1.0 h and about 5% low at 3.0 h. The results for the fourth DR4 ase are essentially the same as those for DR4 ase 3. Thus, one may improve the DR4 alulation with the appropriate hoie of input parameters, inreasing ibrt being a hoie of last resort. Unfortunately, the fluene must be omputed at eight or more loations before the interpolated fluene an be alulated at the detetor loation of interest. Finally, the test program results are in good agreement with the GENMASH results, being about 7% high at 1.0 h and about 2% low at 3.0 h. 16

32 4.0 ONLUSIONS An approximate method (DR4 idos=3) for omputing fluenes, doses, and protetion fators in strutures from plume soures in air-over-ground environments has been developed and tested. This method and a more rigorous method are desribed, and alulated results from eah are ompared with eah other and with free-field doses alulated by GENMASH. ompared to the GENMASH point kernel method, the approximate method yields free-field dose rates that are in good agreement. In addition, sine time- and energy-dependent diretional fluenes an be approximated in the viinity of the struture, one an also estimate shielded fluene spetra and dose rates inside the struture. A great omputational time redution is realized when the approximate method rather than the rigorous method is used, and protetion fators appear to be in reasonably good agreement when suffiient adjoint histories are run in the MASH-MORSE alulations. 17

33

34 5.0 REFERENES 1. G. E. Hansen and H. A. Sandmeier, "Neutron Penetration Fators Obtained by Using Adjoint Transport alulations," Nul. Si. Eng. 22, (July 1965). 2. T. J. Hoffmann, J.. Robinson, and P. N. Stevens, "The Adjoint Differene Method and Its Appliation to Deep-Penetration Radiation Transport," Nul. Si. Eng. 48, (June 1972). 3.. O. Slater and J.. Robinson, "Forward-Adjoint oupling as a Means of Solving Three-Dimensional Deep-Penetration Transport Problems," Nul. Si. Eng. 53, (Marh 1974). 4. W. A. Rhoades, M. B. Emmett, G. W. Morrison, J. V. Pae, III, and L. M. Petrie, "Vehile ode System (VS) User's Manual," Oak Ridge National Laboratory, ORNL/TM-4648 (August 1974). 5. J. O. Johnson, editor, "A User's Manual for MASH A Monte arlo Adjoint Shielding ode," Oak Ridge National Laboratory, ORNL/TM (Marh 1992). 6.. O. Slater, "DR2: A ode with Speialized Appliations for oupling Loalized Monte arlo alulations with Fluenes from Two-Dimensional Disrete Ordinates Air-Over-Ground alulations," Oak Ridge National Laboratory, ORNL/TM (January 1992). 7. J. O. Johnson, editor, "A User's Manual for MASH v1.5 - A Monte arlo Adjoint Shielding ode," Oak Ridge National Laboratory, ORNL/TM-11778/R1 (Otober 1998). 8. J. O. Johnson, editor, "A User's Manual for MASH v2.0 - A Monte arlo Adjoint Shielding ode," Oak Ridge National Laboratory, ORNL/TM-11778/R2 (1999). 9. Defense Speial Weapons Ageny, HPA Version 3.1 Users Guide, Referene Manual Version 7.5 (1996). 10. J. P. Renier, The GENMASH ode is an undoumented addition to the HASAL/SIPUFF ode system. 11. W. A. Rhoades and R. L. hilds, "The TORT Three-Dimensional Disrete Ordinates Neutron/Photon Transport ode," Oak Ridge National Laboratory, ORNL-6268 (November 1987) O. Slater, The GENTORT ode is doumented in Appendix A. 19

35 13. W. A. Rhoades and D. B. Simpson, "The TORSET Method for onstrution of TORT Boundary Soures from External TORT Fluene Files," Oak Ridge National Laboratory (Draft Doument, June 1996), in "DOORS3.1: One-, Two- and Three-Dimensional Disrete Ordinates Neutron/Photon Transport ode System," RSI omputer ode olletion, -650 (August 1996) O. Slater, The PLATEQ ode is doumented in Appendix A O. Slater, The VIST3DP ode is doumented in Appendix A O. Slater, The DR3SRF ode is doumented in Appendix A. 17. D. T. Ingersoll, R. W. Roussin,. Y. Fu, and J. E. White, "DABL69: A Broad-Group Neutron/Photon ross-setion Library for Defense Nulear Appliations," ORNL/TM (June 1989). 18. The International ommittee on Radiologial Protetion, "IRP Publiation 51," Pergamon Press, New York (1987). 19. Amerian Nulear Soiety Standards ommittee, "Neutron and Gamma-Ray ross Setions for Nulear Radiation Protetion alulations for Nulear Power Plants," ANSI/ANS-6.1.2, Amerian Nulear Soiety, La Grange Park, IL, rev. (1999) V. Parks, ed., Sale: A Modular ode System for Performing Standardized omputer Analyses for Liensing Evaluation, NUREG/R-0200 (ORNL/NUREG/SD-2/V2/R1), N. M. Greene, W. E. Ford III, L. M. Petrie, and J. W. Arwood, AMPX: A Modular ode System for Generating oupled Multigroup Neutron-Gamma ross-setion Libraries from ENDF/B-IV and/or ENDF/B-V, ORNL/SD/TM-283 (Otober 1992) O. Slater, The QUAD3D ode is doumented in Appendix A. 23. W. W. Engle Jr., "ANISN, A One-Dimensional Disrete Ordinates Transport ode With Anisotropi Sattering," Report K-1693 (Marh, 1967) O. Slater, The G2GAIR ode is doumented in Appendix A. 20

36 !1!1 Table 1. Atomi Densities m ) of Elements in Mixtures Used in the Test alulations. a b Element a Mixtures Ground Portland Rebar onrete Steel Air b onrete onrete Bloks e H 4.010! ! ! ! ! ! ! ! ! !3 N 4.896! !5 O 4.127! ! ! ! !5 Na 1.470! ! ! !4 Mg 9.403! ! ! !4 Al 1.418! ! ! !4 Si 9.803! ! ! !3 S 3.720!6 l 3.645!6 Ar 2.368!7 K 2.110! ! ! !5 a 2.678! ! ! !3 r 3.205! !3 Mn 7.418! ! !4 Fe 3.310! ! ! ! !2 o 2.867!7 Ni 2.878!7 u 5.006!7 Sn 7.537!8 32% moisture (from ORNL/TM-12685). 0.9 Portland onrete steel by volume Portland onrete. d 21

37 d Air #2 mixture from ORNL/TM Read as e!2 22

38 Table 1. (ontinued) Element Mixtures Rebar Ground Steel Air 2 b Doors Windows onrete 2 a d H 7.768!3 9.75! ! ! !4 N 4.19!5 O 4.386!2 3.48!2 1.13! !4 Na 1.336!3 Mg 1.484!6 Al 2.389!3 4.88!3 Si 1.582!2 1.16! ! !4 S l Ar 2.51!7 K 6.931!4 a 2.915!3 r 1.168!4 Mn 3.878! !6 Fe 4.513! ! !3 o Ni u Sn a b onrete (omposition not shown) + 5% 1020 steel steel (1.111-m-thik steel plate smeared over a m thikness) SiO (0.6-m-thik glass pane smeared over a m-thikness). 2 23

39 d!3 Read as

40 Table 2. Extent of Strutures in the Geometry for TORT/DR3SRF Test Problem 1. oordinate Boundary System Ground Building Plume a XMIN! ! ! XMAX ! YMIN! ! ! YMAX ZMIN!50.0! ZMAX a 12 3 Plume Volume = m. 25

41 Table 3. omparison of Dose Rates for Forward-Forward (TORT-TORT) and Forward-Adjoint (TORT-MASH) ouplings. a b a b Position Soure TORT-TORT (TT) TORT-MASH (TM) f.s.d. TM/TT 1 A 1.79! ! B 1.79! ! A 5.87! ! B 5.88! ! A 3.50! ! B 3.43! ! A 1.71! ! B 1.72! ! A 3.92! ! B 3.80! ! A 2.11! ! B 2.10! ! A 7.77! ! B 7.72! ! A 4.48! ! B 3.64! ! A 9.25! ! B 8.30! ! A 6.08! ! B 5.27! ! (A): Plume soure plus uniform isotropi soure on roof. (B): Plume soure only. Frational standard deviation of the TM result. Read as !19 26

42 Table 4. Energy Boundaries for the Photon Group Strutures Used in the alulations. Group Energy Boundaries (ev) Group a b MASH-MORSE GENMASH TORT d e e e a 17 b 18 DABL69 46n/23g group struture. AMPX Standard 18-group photon struture. 27

43 19 d 7 Struture for SALE 18-group photon library. Read as e Lower energy boundary for the group struture. 28

44 Table 5. ANSI Standard Photon Dose Rates (rem/h) at Detetor Loations Inside a Two-Story onrete Building. Detetor Time (h) a !2 1.47!2 1.40!2 1.30! !3 3.98!3 3.76!3 3.46! !1 5.77!1 6.37!1 1.11!3 1.03!3 9.72!4 8.95! !3 4.93!3 4.66!3 4.31! ! !3 1.88!3 1.78!3 1.65! !2 1.51!2 1.44!2 1.34! !3 5.37!3 5.07!3 4.67! !3 2.67!3 2.52!3 2.32! !3 6.42!3 6.08!3 5.60! !3 3.89!3 3.68!3 3.40!3 a 0 Read as

45 Table 6. ANSI Standard Free-Field Photon Dose Rates (rem/h) at Detetor Loations Inside a Two-Story onrete Building. Detetor Time (h) a !2 7.45!2 7.17!2 6.74! !2 7.03!2 6.76!2 6.36! !2 4.83!2 4.65!2 4.38! !2 7.16!2 6.88!2 6.48! !2 4.86!2 4.67!2 4.40! !1 1.12!1 1.07!1 1.01! !1 1.12!1 1.07!1 1.00! !1 1.14!1 1.08!1 1.02! !1 1.13!1 1.08!1 1.02! !1 1.13!1 1.09!1 1.02!1 a 1 Read as

46 Table 7. ANSI Standard Photon Dose-Rate Protetion Fators at Detetor Loations Inside a Two-Story onrete Building. Detetor Time (h) a a 0 Read as

47 Table 8. omparisons of TORT and GENMASH Free-Field Dose Rates at X= km, Y=55 m, and Z= m. Time After Aident (h) Gamma-Ray Dose Rate (rem/h) GENMASH TORT TORT GENMASH &

48 Table 9. TORT/DR3SRF Free-Field Fluene and Dose Rate and Protetion Fators at a orner Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (oupled neutron and photon adjoint). Time (h) Quantity Free-Field Values a b Fluene Dose ! !2 Dose ! !2 Dose ! !2 Dose ! !2 Protetion Fators For Ground, Plume, and Roof Soures a Fluene Dose Dose Dose Dose Protetion Fators for Ground and Plume Soures a Fluene Dose Dose Dose Dose a!2!1 s. b 7 Read as Doses 1-4 are respetively Straker-Morrison, Henderson, laiborne-trubey, and ANSI standard with units of rem/h

49 Table 10. TORT/DR3SRF Free-Field Fluene and Dose Rate and Protetion Fators at a enter Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (oupled neutron and photon adjoint). Time (h) Quantity Free-Field Values a b Fluene Dose ! !2 Dose ! !2 Dose ! !2 Dose ! !2 Protetion Fators For Ground, Plume, and Roof Soures a Fluene Dose Dose Dose Dose Protetion Fators for Ground and Plume Soures a Fluene Dose Dose Dose Dose a!2!1 s. b 7 Read as Doses 1-4 are respetively Straker-Morrison, Henderson, laiborne-trubey, and ANSI standard with units of rem/h

50 Table 11. TORT/DR3SRF Free-Field Fluene and Dose Rate and Protetion Fators at a orner Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (Photon only adjoint). Time (h) Quantity Free-Field Values a b Fluene Dose ! !2 Dose ! !2 Dose ! !2 Dose ! !2 Protetion Fators For Ground, Plume, and Roof Soures a Fluene Dose Dose Dose Dose Protetion Fators for Ground and Plume Soures a Fluene Dose Dose Dose Dose a!2!1 s. b 7 Read as Doses 1-4 are respetively Straker-Morrison, Henderson, laiborne-trubey, and ANSI standard with units of rem/h

51 Table 12. Free-Field Fluene and Dose Rate and Protetion Fators at a enter Detetor Position in a Large, Thin-Walled Building for a Plume Photon Soure at Several Time Intervals (Photon only adjoint). Time (h) Quantity Free-Field Values a b Fluene Dose ! !2 Dose ! !2 Dose ! !2 Dose ! !2 Protetion Fators For Ground, Plume, and Roof Soures a Fluene Dose Dose Dose Dose Protetion Fators for Ground and Plume Soures a Fluene Dose Dose Dose Dose a!2!1 s. b 7 Read as Doses 1-4 are respetively Straker-Morrison, Henderson, laiborne-trubey, and ANSI standard with units of rem/h

52 Table 13. Desriptions of the 3-D Quadrature Sets on the File reated by QUAD3D. Set Number Set Name Number of Diretions Desription 1 s8 96 symmetri S 8 2 s symmetri S 10 3 s symmetri S 12 4 s symmetri S 14 5 s symmetri S 16 6 d downward biased set (130 down) 7 u upward biased set (130 up) 8 d downward biased set (262 down) 9 u upward biased set (262 up) 10 d downward biased set (306 down) 11 u upward biased set (306 up) 12 d downward biased set (560 down) 13 u upward biased set (560 up) 37

53 Table 14. Free-Field Data and Protetion Fators for a Detetor at Position 1 in a Two-Story Building. The DR4 alulation used a 22-angle 1-D quadrature and an S 8 3-D quadrature to bin the 3-D fluenes at eight fluene points surrounding the building for eah time step (alulation time ~23 min. for two time steps). a a Time Step Quantity DR3SRF DR4 DR3SRF/DR4 Free-Field Data b 1 Fluene Dose Dose Dose Dose Fluene Dose Dose Dose Dose Protetion Fators 1 Fluene Dose Dose Dose Dose Fluene Dose Dose Dose Dose Doses 1-4 are respetively Straker-Morrison, Henderson, laiborne-trubey, and ANSI standard with 38

54 19!2!1 units of rem/h. The fluene has units of s. b 7 Read as

55 Table 15. Free-Field Data and Protetion Fators for a Detetor at Position 2 in a Two-Story Building. The DR4 alulation used a 22-angle 1-D quadrature and an S 8 3-D quadrature to bin the 3-D fluenes at eight fluene points surrounding the building for eah time step (alulation time 1.2 min. for two time steps beause of time savings from the Table 1 alulations). a a Time Step Quantity DR3SRF DR4 DR3SRF/DR4 Free-Field Data b 1 Fluene Dose Dose Dose Dose Fluene Dose Dose Dose Dose Protetion Fators 1 Fluene Dose Dose Dose Dose Fluene Dose Dose Dose Dose Doses 1-4 are respetively Straker-Morrison, Henderson, laiborne-trubey, and ANSI standard with 40

56 19!2!1 units of rem/h. The fluene has units of s. b 7 Read as

57 !2!1 Table 16. Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a two-story a a onrete building (ibrt=2 ; nrad=43 ; DABL69 23-group photon struture; time=38.8 min). Time (h) Quantity a Fluenes and Dose Rates b Fluene Dose Dose Dose Dose Dose Protetion Fators for Fluenes and Dose Rates Fluene Dose Dose Dose Dose Dose ibrt is the number of fluene points per side of a box surrounding the region of interest and nrad is the number of radial distanes at whih ANISN air-attenuated angular fluenes are given). Read as Doses 1-5 are respetively Straker-Morrison, Henderson, laiborne-trubey, ANSI standard, and the AP photon dose rate in a phantom. 20 b

58 !2!1 Table 17. Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a two-story a a onrete building (ibrt=2 ; nrad=82 ; SALE 18-group photon struture; time=34.2 min). Time (h) Quantity a Fluenes and Dose Rates b Fluene Dose Dose Dose Dose Dose Protetion Fators for Fluenes and Dose Rates Fluene Dose Dose Dose Dose Dose ibrt is the number of fluene points per side of a box surrounding the region of interest and nrad is the number of radial distanes at whih ANISN air-attenuated angular fluenes are given). Read as Doses 1-5 are respetively Straker-Morrison, Henderson, laiborne-trubey, ANSI standard, and the AP photon dose rate in a phantom. 20 b

59 !2!1 Table 18. Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a two-story a a onrete building (ibrt=2 ; nrad=43 ; SALE 18-group photon struture; time=34.1 min). Time (h) Quantity a Fluenes and Dose Rates b Fluene Dose Dose Dose Dose Dose Protetion Fators for Fluenes and Dose Rates Fluene Dose Dose Dose Dose Dose ibrt is the number of fluene points per side of a box surrounding the region of interest and nrad is the number of radial distanes at whih ANISN air-attenuated angular fluenes are given). Read as Doses 1-5 are respetively Straker-Morrison, Henderson, laiborne-trubey, ANSI standard, and the AP photon dose rate in a phantom. 20 b

60 !2!1 Table 19. Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a two-story a a onrete building (ibrt=3 ; nrad=43 ; DABL69 23-group photon struture; time=132.5 min). Time (h) Quantity a Fluenes and Dose Rates b Fluene Dose Dose Dose Dose Dose Protetion Fators for Fluenes and Dose Rates Fluene Dose Dose Dose Dose Dose ibrt is the number of fluene points per side of a box surrounding the region of interest and nrad is the number of radial distanes at whih ANISN air-attenuated angular fluenes are given). Read as Doses 1-5 are respetively Straker-Morrison, Henderson, laiborne-trubey, ANSI standard, and the AP photon dose rate in a phantom. 20 b

61 !2!1 Table 20. Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a two-story a a onrete building (ibrt=3 ; nrad=82 ; sale 18-group photon struture; time=116.2 min). Time (h) Quantity a Fluenes and Dose Rates b Fluene Dose Dose Dose Dose Dose Protetion Fators for Fluenes and Dose Rates Fluene Dose Dose Dose Dose Dose ibrt is the number of fluene points per side of a box surrounding the region of interest and nrad is the number of radial distanes at whih ANISN air-attenuated angular fluenes are given). Read as Doses 1-5 are respetively Straker-Morrison, Henderson, laiborne-trubey, ANSI standard, and the AP photon dose rate in a phantom. 20 b

62 !2!1 Table 21. Free-field fluenes s ) and dose rates (rem/h) for detetor 1 in a two-story a a onrete building (ibrt=3 ; nrad=43 ; sale 18-group photon struture; time=116.0 min). Time (h) Quantity a Fluenes and Dose Rates b Fluene Dose Dose Dose Dose Dose Protetion Fators for Fluenes and Dose Rates Fluene Dose Dose Dose Dose Dose ibrt is the number of fluene points per side of a box surrounding the region of interest and nrad is the number of radial distanes at whih ANISN air-attenuated angular fluenes are given). Read as Doses 1-5 are respetively Straker-Morrison, Henderson, laiborne-trubey, ANSI standard, and the AP photon dose rate in a phantom. 20 b

63 Table 22. omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for an Off-enter Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by a 23-group photon library; attenuation data given for 43 distanes). Time (h) TORT-DR3SRF DR4 DR4/TORT-DR3SRF Free-Field Fluenes a Shielded Fluenes Protetion Fators a 7 Read as

64 Table 23. omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for a entered Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by a 23-group photon library; attenuation data given for 43 distanes). Time (h) TORT-DR3SRF DR4 DR4/TORT-DR3SRF Free-Field Fluenes a Shielded Fluenes Protetion Fators a 7 Read as

65 Table 24. omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for an Off-enter Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 82 distanes). Time (h) TORT-DR3SRF DR4 DR4/TORT-DR3SRF Free-Field Fluenes a Shielded Fluenes Protetion Fators a 7 Read as

66 Table 25. omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for a entered Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 82 distanes). Time (h) TORT-DR3SRF DR4 DR4/TORT-DR3SRF Free-Field Fluenes a Shielded Fluenes Protetion Fators a 7 Read as

67 Table 26. omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for an Off-enter Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 43 distanes). Time (h) TORT-DR3SRF DR4 DR4/TORT-DR3SRF Free-Field Fluenes a Shielded Fluenes Protetion Fators a 7 Read as

68 Table 27. omparison of TORT-DR3SRF and DR4 Fluenes and Protetion Fators for a entered Detetor in the Ten-Story NGI Building D (inludes roof soure ontribution; air attenuation by an 18-group photon library; attenuation data given for 43 distanes). Time (h) TORT-DR3SRF DR4 DR4/TORT-DR3SRF Free-Field Fluenes a Shielded Fluenes Protetion Fators a 7 Read as

69 Table 28. omparison of fluenes and dose rates alulated by speial odes using GENMASH soures, the DR4 soure-integration proedure, and ANISN air-attenuated angular fluenes in two group strutures. a Time (h) a a Detetor 1 Detetor 2 b b b b Grp.-A Grp.-B Grp.-B/Grp.-A Grp.-A Grp.-B Grp.-B/Grp.-A Fluene Straker-Morrison Dose Rate (rem/h) Detetor 1 is off enterline and is loated athe oordinates ( e5, 30.48, ) relative to the GENMASH soure mesh. The on-enterline detetor 2 is loated at the oordinates (-8.0e5, 0.0, ). The oordinate unit is m. b Grp.-A is the bottom 15 groups of the SALE 18-group photon struture and Grp.-B is the bottom 16 groups of the DABL69 23-group photon struture. 7 Read as

70 Table 29. omparison of GENMASH dose rates with TORT, DR4, and Test Program dose rates. Time (h) Method Dose Rate (rem/h) % Diff. Dose Rate (rem/h) % Diff. a b d GENMASH TORT 24.73! !40.4 a DR ! !25.0 b DR ! !9.23 DR !4.70 d DR !4.77 Test Program !1.61 ibrt=2, xmin=0, xmax=9000, ymin=0, ymax=6000, zmin=0, zmax=800, xd=1250, yd=5500, zd=770.04, xo(i)=!8.1e5, yo(i)=0, zo(i)=0. ibrt=3, xmin=0, xmax=9000, ymin=0, ymax=6000, zmin=0, zmax=800, xd=1250, yd=5500, zd=770.04, xo(i)=!8.1e5, yo(i)=0, zo(i)=0. ibrt=2, xmin=0, xmax=500, ymin=0, ymax=500, zmin=0, zmax=800, xd=250, yd=250, zd=770.04, xo(i)=!8.085e5, yo(i)=5250, zo(i)=0. ibrt=2, xmin=0, xmax=500, ymin=0, ymax=500, zmin=0, zmax=500, xd=250, yd=250, zd=250, xo(i)=!8.085e5, yo(i)=5250, zo(i)=

71 56

72 Figure 1. Exterior view of the two-story onrete building used in the DR3SRF and DR4 test ases. 57

73 Figure 2. Exterior view of the roofless two-story onrete building used in the DR3SRF and DR4 test ases. 58

74 Figure 3. X-Y planar slie through building to show X-Y loations of detetors within the building. The first- and seond-floor detetors have the same X and Y oordinates. 59

75 !3!1 Figure 4. Surfae plot of the energy-integrated soure s ) at 1-m height one hour after the start of a PWR aident. 60

76 !3!1 Figure 5. Surfae plot of the energy-integrated soure s ) at 1-m height three hours after the start of a PWR aident. 61