ADVANCED NEUTRON IRRADIATION SYSTEM USING TEXAS A&M UNIVERSITY NUCLEAR SCIENCE CENTER REACTOR

Transcription

1 ADVANED NEUTRON IRRADIATION SYSTEM USING TEXAS A&M UNIVERSITY NULEAR SIENE ENTER REATOR A Dissertation by SI YOUNG JANG Submitted to the Offie of Graduate Studies of Texas A&M University in partial fulfillment of the requirements for the degree of DOTOR OF PHILOSOPHY August 2004 Major Subjet: Nulear Engineering

2 ADVANED NEUTRON IRRADIATION SYSTEM USING TEXAS A&M UNIVERSITY NULEAR SIENE ENTER REATOR A Dissertation by SI YOUNG JANG Submitted to Texas A&M University in partial fulfillment of the requirements for the degree of DOTOR OF PHILOSOPHY Approved as to style and ontent by: Warren D. Reee (hair of ommittee) Mihael A. Walker (Member) Leslie A. Braby (Member) John Ford (Member) William E. Burhill (Head of Department) August 2004 Major Subjet: Nulear Engineering

3 iii ABSTRAT Advaned Neutron Irradiation System Using Texas A&M University Nulear Siene enter Reator. (August 2004) Si Young Jang, B.S., Hanyang University; M.S., Korea Advaned Institute of Siene and Tehnology hair of Advisory ommittee: Dr. Warren D. Reee A heavily filtered fast neutron irradiation system (FNIS) was developed for a variety of appliations, inluding the study of long-term health effets of fast neutrons by evaluating the biologial mehanisms of damage in ultured ells and living animals suh as rats or mie. This irradiation system inludes an exposure ave made with a lead-bismuth alloy, a ave positioning system, a gamma and neutron monitoring system, a sample transfer system, and interhangeable filters. This system was installed in the irradiation ell of the Texas A&M University Nulear Siene enter Reator (NSR). By inreasing the thikness of the lead-bismuth alloy, the neutron spetra were shifted into lower energies by the sattering interations of fast neutrons with the alloy. It is possible, therefore, by hanging the alloy thikness, to produe distintly different dose weighted neutron spetra inside the exposure ave of the FNIS. The alulated neutron spetra showed lose agreement with the results of ativation foil measurements, unfolded by SAND-II lose to the ell window. However, there was a onsiderable less agreement for loations far away from the ell window. Even though the magnitude of

4 iv values suh as neutron flux and tissue kerma rates in air differed, the weighted average neutron energies showed lose agreement between the MNP and SAND-II sine the normalized neutron spetra were in a good agreement eah other. A paired ion hamber system was onstruted, one with a tissue equivalent plasti (A-150) and propane gas for total dose monitoring, and another with graphite and argon for photon dose monitoring. Using the pair of detetors, the neutron to gamma ratio an be inferred. With the 20 m-thik FNIS, the absorbed dose rates of neutrons measured with the paired ion hamber method and alulated with the SAND-II results were 13.7 ± 0.02 Gy/min and 15.5 Gy/min, respetively. The absorbed dose rate of photons and the gamma ontribution to total dose were ± Gy/min and 4.7%, respetively. However, the estimated gamma ontribution to total dose varied between 3.6 % to 6.6 % as the assumed neutron sensitivity to the graphite detetor was hanged from 0.01 to 0.03.

5 v AKNOWLEDGMENTS I would like to forever aknowledge that none of this work would be possible without Jesus hrist. First of all, I would like to thank my advisor, Dr. Dan Reee, for his sinere guidane and support in my graduate eduation, as well as livelihood in ollege Station. In addition, I would like to express my thanks to Dr. Leslie Braby for his endless and sinere tehnial advie. Speial thanks go to Dr. John Ford and Dr. Mihael Walker for their help and support as members of my ommittee. I would like to express my thanks to the NS staff for their help and friendship. Finally, I would like to thank my family, my wife Kiyeon, my daughter Jiwon, and my son Sungjin, for their tremendous support.

6 vi TABLE OF ONTENTS Page ABSTRAT... AKNOWLEDGMENTS... TABLE OF ONTENTS... LIST OF FIGURES... LIST OF TABLES... iii v vi viii xiv HAPTER I INTRODUTION Bakground Literature Review Objetive... 4 II MATERIALS AND METHODS Overview of Nulear Siene enter Reator General Overview NSR ore harateristis Irradiation ell Filtered Neutron Irradiation System (FNIS) Design Requirements Properties of Lead-Bismuth Alloy and Boral Plate Exposure ave Sample Transfer and ontrol System Radiation Monitoring System Paired Ion hamber Method Stainless Steel-Walled Ion hamber Ativation Foil Method for Determining Neutron Spetra Monte arlo Modeling Overview Reator ore Modeling Irradiation ell Modeling Inluding the FNIS... 49

7 vii HAPTER Page ritiality and Fluene alulation Absorbed Dose Rate alulation III RESULTS AND DISUSSION Sensitivity Analysis Results Impat of Lead-Bismuth Thikness on the Neutron Energy Spetra Impat of Boral Box on the Neutron Energy Spetra Impat of Aluminum and Other Materials on the Neutron Energy Spetra Impat of Reator Gap on the Neutron Energy Spetra omparison of MNP Results with Experiment Data omparison of MNP Results with Experimental Data Without the FNIS omparison of MNP Results with Experimental Data for the FNIS (I) omparison of MNP Results with Experimental Data for the FNIS (II) Axial and Radial Distribution of Neutrons in the FNIS omparison of Tissue Kerma Rate in Air with Absorbed Dose Rate in Water IV ONLUSION REFERENES APPENDIX A MONTE ARLO SOURE PROGRAMS VITA

8 viii LIST OF FIGURES Page Fig. 2.1 Fig. 2.2 Fig. 2.3 a) General view of the NSR at stall position. b) Upper researh level plan of the NSR, whih shows a huge irradiation ell, reator pool, and ontrol room. ) ross-setion showing reator positioned adjaent to entrane window of dry well. 8 NSR reator ore onfiguration inluding ontrol rods, graphite, and ore monitoring instruments. 11 NSR general drawing inluding experimental failities and irradiation ell.. 13 Fig. 2.4 Top and side views of the irradiation ell at the NSR Fig. 2.5 Lead-bismuth ingots, whih ontains 44.5 w/o lead (Pb) and 54.9 w/o bismuth (Bi), and lead-bismuth briks for the FNIS setup, whih have the size of m m 5.08 m.. 17 Fig. 2.6 Mass attenuation oeffiients of lead and bismuth, respetively Fig. 2.7 Total (MT=1) and sattering (MT=2) ross-setion data of lead and bismuth, respetively Fig. 2.8 Boral box onstruted for shielding against thermal neutrons Fig. 2.9 Fig Fig Fig Exposure ave for sample irradiation, whih was onstruted with the lead-bismuth briks 24 omponents used to set up the exposure ave inside the irradiation ell, and final setup of the FNIS inluding the exposure ave, movable working table, trak, and detetors installed inside the irradiation ell Diagram of the sample transfer system installed in the upper researh level (top) and in the irradiation ell (bottom) for sample loading/unloading and exposure ontrol 27 Home-made sensor made of miro-swith (left-hand side) and LED/photo-transistor (right-hand side) for blower ontrol, as well as for exposure time ontrol 28

9 ix Page Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Sample transfer system installed in the upper researh level and in the irradiation ell for sample loading/unloading and exposure ontrol Boron-ontaining polyethylene panel and portable wooden box for minimizing neutron streaming through the penetration and the onrete door.. 29 Sample transfer box installed in the upper researh level for sample loading and unloading. 29 omputerized ontrol system for manipulating eah module of the FNIS system automatially 30 Digital data interfae ard (PI-DIO24 TM manufatured by Measurement omputing orporation), whih was used for data olletion and blower ontrol Design sheme and top/side views of the tissue equivalent and graphite-walled ion hambers used in this study. 33 X-ray mahine (Norelo MG300) of the Texas A&M Nulear Siene enter used for determining operating voltage, whih set at 250 kvp and 10 ma Keithley TM 485 Auto-ranging Pio-ammeter, Hewlett Pakard TM HP6516A D Power Supply used in this study Gas-filling system used in this study for filling argon and propane gas into graphite and tissue equivalent plasti ion hambers, respetively.. 37 Measured urrent with varying voltages for setting the operating voltages of the detetors used in this study Neutron ativation flux foils, whih onsist of admium-overed foils suh as Al, Fe, Au, Ni, Ti, Zn, In, o, Mg, and u and a bare aluminum gold foil. Eah foil paket is approximately 4 4 m 41

10 x Page Fig Fig MNP drawings denoting NSR, void box, ell window, iradiation ell, and the exposure ave on x-y plane, y-z plane, and x-z plane 45 Fuel element and fuel bundle as modeled in this study. Note that the fuel meat setion is divided into 5 segments and this study alulates neutron flux in eah of these five segments.. 46 Fig Monte arlo modeling of ontrol rods as modeled in this study. 47 Fig Fig. 3.1 Fig. 3.2 Fig. 3.3 Fig. 3.4 Fig. 3.5 Fig. 3.6 Fig. 3.7 MNP drawings denoting exposure ave surrounded by a boral box, a movable steel table, a ell window, and the walls of the irradiation ell on a) x-y plane, b) y-z plane, and ) x-z plane 50 Differential neutron flux over all neutron energies with varying thikness of lead-bismuth alloy, inluding a void box, a boral plate, and a boral box. 56 Differential neutron flux for fast neutron energies with varying thikness of lead-bismuth alloy, inluding a void box, a boral plate, and a boral box. 56 Fration of neutron flux over all neutron energies with varying thikness of lead-bismuth alloy, inluding a void box, a boral plate, and a boral box Fration of tissue kerma in air of neutrons with varying thikness of lead-bismuth alloy for fast neutron energies, inluding a void box, a boral plate, and a boral box Mean neutron energies of tissue kerma rate in air and of neutron flux with inreasing thikness of lead-bismuth alloy in the exposure ave ontribution of gamma rays to total dose with varying thikness of lead-bismuth alloy in the exposure ave Tissue kerma rate in air of neutrons and of photons with varying thikness of lead-bismuth alloy, respetively.. 61

11 xi Page Fig. 3.8 Fig. 3.9 Fig Fig Fig Fig Fig Fig Fig Fig Tissue kerma rate in air of neutrons inside the exposure ave with varying thikness of lead-bismuth alloy, whih was alulated with different kerma fators stated in the IRU publiation Differential neutron energy spetra simulated by the MNP with and without boral box, respetively, with a given 10 m-thik lead-bismuth alloy over all neutron energies Fration of neutron flux over all neutron energies with and without boral box, respetively, with a given 10 m-thik lead-bismuth alloy Fration of tissue kerma in air over all neutron energies with and without boral box, respetively, with a given 10 m-thik lead-bismuth alloy Differential neutron energy spetra for fast neutron energies simulated by the MNP with varying thikness of aluminum plate Fration of neutron flux over all neutron energies with varying thikness of aluminum plate for a given no lead-bismuth alloy in the FNIS Fration of tissue kerma rate in air of neutrons for fast neutron energies with varying thikness of aluminum plate for a given no lead-bismuth alloy in the FNIS 69 Mean neutron energies weighted by neutron flux and by tissue kerma rate in air of neutrons with inreasing thikness of aluminum plate 70 ontribution of gamma rays to total dose with inreasing thikness of aluminum plate 70 Mean neutron energies for the tissue kerma rate in air of neutrons and neutron flux with variable sizes of reator gap installed between the void box and the boral plate in the reator pool-side, respetively. 73

12 xii Page Fig Fig Fig Fig Fig Fig Fig Fig Fig Tissue kerma rate in air of neutrons and of photons with variable sizes of reator gap installed between the void box and the boral plate in the pool-side, respetively Differential neutron energy spetra simulated by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils over all neutron energies without FNIS.. 78 Differential neutron energy spetra simulated by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils for fast neutron energies without FNIS The ratio of the MNP to SAND-II results without the FNIS in the irradiation ell for the distane from the irradiation ell window.. 79 Tissue kerma rate in air of neutrons simulated by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils without FNIS.. 79 Mean neutron energies based on the MNP and SAND-II results of the tissue kerma rate in air of neutrons and neutron flux without FNIS, respetively Normalized neutron spetra from SAND-II and MNP at the distane of 183 m away from the irradiation ell window without the FNIS (The alulated spetrum was normalized so that the total neutron flux within the energy range of MeV was equal to one) Differential neutron energy spetra obtained by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils over all neutron energies with the FNIS, onstruted of 10 m thik lead-bismuth alloy Differential neutron energy spetra obtained by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils for fast neutron energies with the FNIS, onstruted of 10 m thik lead-bismuth alloy 85

13 xiii Page Fig Fig Fig Fig Fig Fig Fig Fig Fig Differential neutron energy spetra alulated by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils with 20 m-thik Pb-Bi alloy over all neutron energies.. 89 Differential neutron energy spetra alulated by the MNP, as well as by the SAND-II using the measured speifi ativities of ativation foils with 20 m-thik lead-bismuth alloy for fast neutron energies 89 omparison of the differential neutron energy spetra obtained by the SAND-II using the measured speifi ativities of ativation foils behind 10 m-thik and 20 m-thik lead-bismuth alloy over all neutron energies omparison of the differential neutron energy spetra obtained by the SAND-II using the measured speifi ativities of ativation foils behind 10 m-thik and 20 m-thik lead-bismuth alloy for fast neutron energies Tissue kerma rate in air of neutrons and of photons measured with paired ionization hamber method and gamma ontribution to total dose in perentage with the FNIS, onstruted of 20 m thik lead-bismuth alloy, respetively. 94 Relative neutron sensitivities of different types of detetors with inreasing neutron energy 94 Radial distribution of the speifi ativities of 115m In produed by the (n, n ) reation of fast neutrons with the indium foils at 5 m distane from the front fae of the exposure ave of 20 m thik lead-bismuth alloy Axial distribution of the speifi ativities of 115m In produed by the (n, n ) reation of fast neutrons with the indium foils from the front fae of the exposure ave of 20 m thik lead-bismuth alloy omparison of absorbed dose rates in tissue and tissue kerma rates in air with varying depth of tissue... 98

14 xiv LIST OF TABLES Page Tabel 2.1 harateristis of the low melting alloy of lead-bismuth.. 16 Tabel 2.2 alibration fator for the paired-ion hambers used in this study Tabel 2.3 harateristis of the flux foils in terms of neutron reation Tabel 3.1 Tabel 3.2 Tabel 3.3 Tabel 3.4 Tabel 3.5 Tabel 3.6 Tabel 3.7 Mean neutron energies weighted by neutron flux and tissue kerma rate in air of neutrons, and gamma ontribution in total dose with various ombinations of neutron filters, inluding a void box, a boral plate, and a boral box Neutron flux alulated using the MNP and the SAND-II at different loations from the ell window without the FNIS, respetively (1m gap + safety shim loation at 80%).. 77 Tissue kerma rate in air of neutrons without the FNIS alulated with neutron flux and kerma fator stated in the IRU Neutron flux, tissue kerma rate in air of neutrons alulated with the neutron flux and the kerma fator stated in the IRU-26, and mean neutron energy for the neutron flux and the tissue kerma rate in air with the FNIS, onstruted of 10 m thik lead-bismuth alloy Neutron flux, tissue kerma rate in air of neutrons alulated with the neutron flux and the kerma fator stated in the IRU-26, and mean neutron energy for the neutron flux and the tissue kerma rate in air with the FNIS, onstruted of 20 m thik lead-bismuth alloy Photon ontribution to total dose, whih is produed by interating neutrons with the detetor filling gases using the MNP ode Physial parameters used for the absorbed dose haraterization in this study

15 1 HAPTER I INTRODUTION BAKGROUND In 1994, a Task Group submitted a report titled Biologial Effetiveness of Neutrons: Researh Needs (ORAU 1994) to the Federal Government s ommittee on Interageny Radiation Researh and Poliy oordination. The onluding observation of this report was that an effetive program foused upon the radiobiology of neutrons requires that adequate laboratory and irradiation failities be available and adequately staffed. This report also pointed out that due to the limitations of human data in terms of statistis and types of exposures where data are available, and the limitations of extrapolating animal data to humans, the long term solution to estimating health risks will depend on a better understanding of the underlying mehanisms. Unfortunately, as university and Department of Energy researh reators have been shut down, the number of failities where the needed neutron irradiation spetra an be found has ontinued to diminish. Furthermore, the biologial signifiane of radiation response in ells that have not been traversed by a harged partile has beome a major question in the area of risk evaluation for high LET radiations. Although a great deal of work, reviewed in the IRRP report (ORAU 1994), has been done on the biologial effetiveness of neutrons in animals and ell ulture systems, very little of this an be used to evaluate the signifiane of a bystander ell effet or to guide development of risk evaluation models. Most of the mehanisti 1 This dissertation follows the style of Health Physis.

16 2 studies were done with mammalian ells grown as a mono-layer on a plasti substrate. This irradiation onfiguration, and the ellular and moleular endpoints that were utilized, limited the results to those expressed by individual irradiated ells. In intat organisms, ells seldom exist in mono-layers, or our in isolation from surrounding ells. In order to understand the mehanisms of damage and repair that are mediated, in part, by a bystander ell effets, it is neessary to ondut experiments with threedimensional tissue systems expressing normal ell growth and differentiation harateristis. A reent emphasis of the Department of Energy Low Dose Researh Initiative has been on the development of experimental systems for studying low dose effets, inluding the bystander effet, in tissues and other three-dimensional systems. Several new approahes have been proposed, and researh is progressing. Eventually it will be neessary to test models derived from this tissue response researh in test animals and in in vivo/in vitro systems. Organ ulture and small animal experiments to test radiation risk models plae stringent demands on the neutron irradiation system used to ondut the researh. The biologial effetiveness of neutrons is thought to be a strong funtion of the neutron energy or other measure of radiation quality. ell and tissue responses also seem to be strongly influened by the environmental onditions suh as temperature and atmosphere omposition. Therefore, to test models of radiation risk requires an irradiation faility that provides two or more well haraterized neutron fields with signifiantly different radiation quality, minimum gamma ray ontamination, adequate dose rates, and ontrolled environmental parameters.

17 LITERATURE REVIEW Previous work at the Armed Fores Radiobiology Researh Institute (AFRRI) has shown that it is possible to ahieve gamma dose rates of less than 10 % of the total dose rate in fission neutron beams by enlosing the sample to be irradiated in a shield onstruted primarily of lead (Prasanna et al. 2002). AFRRI has established a neutron irradiator using a TRIGA reator that delivers a fission neutron field with a 30:1 neutron/photon absorbed dose ratio; the design used lead briks and fission neutron beams inluding thermal neutrons, with the photon ontamination mostly shielded. Hene, the neutron spetrum of the AFRRI faility was lose to that of fission neutron spetrum in the reator ore. The AFRRI faility has a movable lead shielding system, whih an be plaed in front of the reator ore to inrease the neuron to gamma ratio (Redpath et al. 1995). In addition, the exposure room is lined with a gadoliniumadmium shielding materials to minimize the sattered thermal neutrons (Zeman et al. 1988). The exposure ave was built of lead briks, 5 m-thik, and the outside dimensions of the system were 71 m in height and 51 m in width and depth. Additional 15-m lead shielding was plaed in front of the exposure ave (Prasanna et al. 1997). The samples were loaded into the ave manually through the extrator tube, whih penetrated the exposure room wall. The University of Massahusetts Lowell Researh Reator has installed a Fast Neutron Irradiator (FNI) in the pool area of the reator (White and Jirapongmed 2002) to provide a large-volume irradiation loation for the irradiation of eletroni parts, whih

18 4 has a high fast neutron flux with low thermal neutron and gamma flux. One of the design goals was to provide neutron beam that have a 10:1 fast-to-thermal flux ratio and a total gamma dose rate to the sample less than 1 kgy/hr (White and Jirapongmed 2002). On the other hand, researh in Boron Neutron apture Therapy (BNT) are urrently under way in many different plaes, utilizing epithermal neutrons with lowlevels of fast neutrons and gamma-ray ontamination (Ross et al. 1993; Matsumoto 1996; Sakurai and Kobayashi 2000). Although these BNT studies have the same goal in minimizing gamma-ray ontamination, materials with high-sattering ross-setions are used to produe epithermal neutrons by slowing down fast neutrons OBJETIVE The objetive of this study is to develop a heavily filtered fast neutron irradiation system (FNIS) that will be used to evaluate the biologial mehanisms whih lead to long-term health effets from neutron exposure. In addition, the FNIS ould be used to test eletroni parts suh as integrated iruit (I) hips and semiondutors with fast neutrons. This irradiation system utilizes the Texas A&M University Nulear Siene enter Reator (NSR), whih is a 1-MW pool-type MTR-onverted TRIGA reator, to provide an intense fission spetrum neutron soure, arefully designed neutron filters to reate spetra with different biologial effetiveness, a heavily shielded exposure ave to minimize gamma dose, and a pneumati sample transfer system to deliver samples for irradiation and to ontrol dose without interrupting reator operation. To aomplish this, an exposure ave of lead-bismuth alloy, a moving rail system, paired ion hambers

19 5 for dose monitoring, sample transfer system, and interhangeable filters were installed into the irradiation ell of the NSR. This study used the Monte arlo N-Partile (MNP) version 5 ode (Briesmeister 2003) and a set of high-temperature ENDF/B-VI ontinuous neutron ross-setion data for a realisti modeling of the reator, the irradiation ell, and the FNIS. Sensitivity analyses were performed to find the harateristis of the FNIS as a funtion of the thikness of the lead-bismuth alloy. A paired ion hamber system was onstruted with a tissue equivalent plasti (A-150) and propane gas for total dose monitoring and with graphite and argon for gamma dose monitoring. This study, in addition, onfirmed the Monte arlo modeling of the FNIS, as well as the performane of the system by omparing the alulated results with experimental data measured with ativation foils and paired ion hambers. ontents of this dissertation are as follows: materials and methods used in this study are presented in hapter II. Setion 2.1 desribes the harateristis of the Nulear Siene enter Reator (NSR), inluding the explanation of both the reator ore and the irradiation ell. Setion 2.2 deals with the design and onstrution of the filtered neutron irradiation system (FNIS) followed by the design requirements and the harateristis of the onstruting materials. Setion 2.3 explains the sample transfer and ontrol system installed at the NSR. The radiation monitoring system is presented in Setion 2.4, inluding the paired ion hamber method, as well as the foil ativation method for haraterizing neutron energy spetra. The Monte arlo modeling is

20 6 desribed in Setion 2.5. This setion ontains the detail simulation sheme of the FNIS using the MNP ode. In hapter III, results and disussion are presented. Setion 3.1 desribes the results of sensitivity analyses using varying thikness of lead-bismuth alloy, the boral box installed outside of the FNIS, different neutron filters installed in front of the FNIS, and the reator gap between the aluminum window of the irradiation ell and the void box of the reator. omparison of MNP results with experimental data is disussed in Setion 3.2. This setion evaluates the 3-dimensional MNP model without and with the FNIS in the irradiation ell by omparing the alulated MNP results with those of neutron spetrum and dose rate measurements. Setion 3.3 ontains the explanation on the axial and radial distributions of neutrons inside the FNIS. omparison of tissue kerma rate in air with absorbed dose rate in water is shown in Setion 3.4. Finally, onlusions of this study are given in hapter IV.

21 7 HAPTER II MATERIALS AND METHODS 2.1. OVERVIEW OF NULEAR SIENE ENTER REATOR General Overview Sine its initial operation on Deember 18, 1961, the Nulear Siene enter Reator at Texas A&M University (NSR) has been heavily utilized (NS 2003). The NSR, whih is a 1-MW, pool-type TRIGA (Teahing, Researh, and Isotopes, General Atomi) reator, has been in ative use for radioisotope prodution, Instrumental Neutron Ativation Analysis (INAA), and various experiment setups suh as delayed neutron ounting, a real-time neutron radiography, and experiments with beam ports. The NSR has a movable reator ore that an be plaed at any position along the pool enterline. On the west side of the pool is an irradiation ell for large experimental objets as shown in Fig The reator has operated with a full FLIP (Fuel Life Improvement Program) ore or FLIP/Standard mixed ore sine 1973; the urrent onfiguration is based on 4-element bundles with TRIGA-FLIP fuel (NS 2003). Light water irulates through the reator ore by natural onvetion, ating as a neutron refletor and reator oolant at the ore perimeter, and graphite serves as an additional refletor on two sides to minimize neutron leakage. The onrete pool struture, inluding the stainless steel liner and the pool water, provide neutron and gamma ray shielding for the reator. The shielding apaity is for a reator operating at 5 MW, whih is well above the urrent 1 MW maximum operating level (NS 2003). The movable reator bridge allows the operation

22 8 of the reator at any position on the pool enterline, whih runs approximately east to west as shown in Fig a Irradiation ell Reator Bridge b Fig a) General view of the NSR at stall position. b) Upper researh level plan of the NSR, whih shows a huge irradiation ell, reator pool, and ontrol room. ) rosssetion showing reator positioned adjaent to entrane window of dry well.

23 9 Sine the fuel onfiguration in the NSR is suh that the minimum nominal spaing between the fuel rods provides adequate onvetion ooling of reator ore up to 2.0 MW thermal power-rating, this spaing and the extra ooling holes at the orners of the bundle an enhane the ore ooling apability. The inreased depth of pool, in addition, improves the ooling apability of the NSR. A suspension frame supported by a bridge yoke that spans the pool supports a grid blok whih in turn supports the fuel, refletor, ontrol rods, samples and any others in-the reator ore. The pool is 10 m-deep and 5.5 m-wide in the main pool side and experimental penetrations onsist of the thermal olumn, pneumati tubes, beam ports and the irradiation ell window NSR ore harateristis The NSR obtained permission to operate full standard, mixed, or full FLIP TRIGA ores in June 1973 (NS 2003). The mixed ore is liensed to operate at a maximum steady-state power of 1 MW with a maximum pulse reativity insertion of $2.00. In July 1973, the first NSR mixed TRIGA ore, ontaining 35 FLIP and 63 standard elements, went into servie. In July 1975, the maximum permissible pulse reativity insertion was inreased to $2.70. The NSR has operated with two mixed ore loadings ontaining 35 FLIP and 59 FLIP sine initial approval in June Sine the late 1970s, the NSR, however, has been operating with all FLIP fuel.

24 10 The urrent NSR ore onfiguration as shown in Fig. 2.2 ontains 92 regular fuel elements, an instrumented fuel element, 4 shim safety rods, a transient rod and a regulating rod (Kim et al. 2004). The fuel elements and ontrol rods are grouped into 4- rod bundles, whih are positioned and supported by an aluminum grid plate ontaining a 6 by 9 array of holes. The graphite bloks, detetors, pneumati devies and various experiments are also positioned and supported by the holes on the grid plate. The grid loations on the ore are addressed by the olumn (A, B,, D, E and F) and row (1, 2, 3, 4, 5, 6, 7, 8 and 9) of the grid plate. The four positions in eah bundle are speified by NW (northwest), NE (northeast), SW (southwest) and SE (southeast). The grid loations in olumn A (more speifially, A2, A4, A6 and A8) on the grid plate are available for positioning various experiments or irradiation devies. The grid loations D3 and D7 are reserved for experiments that require fast neutrons or higher neutron flux. Pneumati irradiation devies are installed at B1, 2, and D2 for short irradiations. The east fae of the reator ore is used to irradiate large objets.

25 11 N Fig NSR reator ore onfiguration inluding ontrol rods, graphite, and ore monitoring instruments (NS 2003).

26 Irradiation ell The irradiation ell (i.e., dry well) is a shielded struture adjoining the main pool as shown in Fig. 2.3 (NS 2003). The irradiation ell was originally designed to perform reator experiments for large objets suh as large animals and motor/pumps or to serve as pool water storage (Gidden 1996). The ell is now, however, used mainly for the radiation exposure of small biologial samples with the NSR or various targets with the lanthanum gamma soure. An irradiation window, whih is loated in the shield wall, separates the reator pool and the irradiation ell. The reator an operate any desired distane from the ell window for irradiating objets in the ell. The upper 5.2 m-length pool wall is made with standard onrete, while the lower portion of the pool wall is barites onrete and light onrete (NS 2003). The NSR an operate at steady state power levels up to 1 MW with the reator plaed against the irradiation ell at the west end of the main pool. Whenever radiation levels in the upper researh level inrease above 2.5 mr/hr, the ontrol room an ontrol all exess points to the upper and lower researh levels. The irradiation ell is 5.49 m wide by 4.88 m deep by 3.05 m high with 0.6 m thik onrete wall between the irradiation ell and reator pool. The window is overed by an aluminum plate 1.27 m thik on the pool-side and is readily aessible for equipment installation and modifiation. The window to the pool is 0.6 m square at the pool-side of the wall, expanding to 1.2 m square on the irradiation ell side. The reator an operate with the front row of fuel elements unfilled, thus inserting a 7.62 m water gap between reator ore fae and irradiation window, or a void box ontaining air an be installed on

27 13 the front row filling the 7.62 m-gap between reator ore fae and the irradiation ell window. A 0.6 m-thik onrete eiling overs the top of the irradiation ell. A 1.5 m by 1.5 m opening provides the aess to the irradiation ell over the irradiation window. A motor-driven onrete slab rolls on traks to over the aess hole when the reator is in operation as shown in Fig The air from the dry well is vented to the entral exhaust system to remove 41 Ar in the ell produed by the neutron ativation of the air, and is monitored by a gas monitor before release to the environment. Fig NSR general drawing inluding experimental failities and irradiation ell.

28 Fig Top and side views of the irradiation ell at the NSR. 14

29 FILTERED NEUTRON IRRADIATION SYSTEM (FNIS) Design Requirements The objetive of this study is to develop a heavily Filtered Neutron Irradiation System (FNIS) suitable for exposing tissue samples and small animals. This irradiation system would take advantage of the irradiation ell at the NSR. Previous work at the Armed Fores Radiobiology Researh Institute (AFRRI) has shown that it is possible to limit gamma dose rates to less than 10% in fission neutron beams by enlosing the sample being irradiated in a shield onstruted primarily of lead (Redpath et al. 1995). The AFRRI design has established a neutron irradiator with a TRIGA reator to deliver a fission neutron field apable of yielding a 30:1 neutron/photon absorbed dose ratio. The detailed design requirements of this study were 1) to deliver fast neutron beams with mean neutron energy greater than 0.1 MeV with several distintly different energy spetra into the exposure ave of the FNIS, 2) to deliver neutron dose to the samples with photon dose ontamination of less than 5%, 3) to produe uniform dose distribution inside the exposure ave, 4) to be able to ahieve a wide range of dose rates by moving the FNIS/reator ore or by adjusting the reator power, 5) provide a sophistiated ontrol system for sample transfer, whih an transfer living animals or tissues with minimum dose reeived while they are in transit through the unfiltered neutron environment, and preventing damage whih might be done by a sudden stop, and 6) provide neutron and gamma ray dosimetry. For aomplishing the design requirements of the new neutron irradiation system, we used materials that have low absorption ross-setions for fast neutrons, as well as

30 16 suitable sattering ross-setion, so as not to degrade fast neutron beams. In addition, we used thermal/epithermal neutron absorbers to ut off undesirable low energy neutrons. We designed the FNIS using a euteti alloy of 54.9% bismuth and 44.5% lead as the major onstrution material, rather than pure lead, to minimize the prodution of apture gamma rays inside the FNIS Properties of Lead-Bismuth Alloy and Boral Plate The euteti lead-bismuth Alloy (44.5 w/o Pb, 54.9 w/o Bi) has a potential to be used as a oolant for liquid metal fast breeder reator instead of liquid sodium (IAEA 2002). Table 2.1 shows some physial and thermal-physial properties haraterizing this alloy (IAEA 1998). In this study, the low melting alloy of lead-bismuth (Belmont 2562G) manufatured by Belmont Metals In., was used to ast lead-bismuth briks for the system setup. The lead-bismuth ingots as shown in Fig. 2.5 were melted using a furnae and then were poured into a steel form m m 5.08 m to make the leadbismuth briks. Table 2.1. harateristis of the low melting alloy of lead-bismuth (Belmont 1997). Properties Nominal Value Element Abundane (w %) Melting Temperature K Sb 0.25 max Boiling Temperature 1943 K Bi 54.9 Density (g/m 3 ) d 0.25 max Heat apaity 146 (J/kg-K) u 0.10 max Pb 44.5

31 Fig Lead-bismuth ingots, whih ontains 44.5 w % lead (Pb) and 54.9 w % bismuth (Bi), and lead-bismuth briks for the FNIS setup, whih have the size of m m 5.08 m. 17

32 18 Although good photon shielding apability and mehanial strength should be onsidered as basi requirements to hoose materials for onstruting a heavy shielding system, low neutron apture ross-setions are desirable to minimize photon prodution inside the system. Figs. 2.6 and 2.7 show the mass attenuation oeffiients (NIST 1996) and neutron ross-setion data for lead and bismuth (ENDF 1998), respetively. In terms of neutron apture reation with the shielding material, pure bismuth is better than pure lead due to low neutron absorption ross-setion of the bismuth over the lower neutron energies, and bismuth, in addition, has a good shielding apability, whih is almost equal to that of lead in attenuating gamma rays from reator ore and surrounding onrete. Bismuth is easy to mahine into any shape. Pure bismuth, however, is too weak to support a heavy shield that may weigh more than 2 tons. To gain the advantages and to avoid the disadvantages of the pure bismuth, a low melting alloy of lead-bismuth was used in this study. In addition, the fission neutron spetrum is usually shifted down in energy whenever the neutrons pass through any high-z materials suh as lead and bismuth. This shift in energy is due to the energy dependene of the inelasti reations, and neutrons below 1 MeV are aumulated as the higher-energy neutrons are sattered into a lower energy region (Prasanna et al. 2002).

33 Fig Mass attenuation oeffiients of lead and bismuth, respetively (NIST 1996). 19

34 20 Pb Bi Fig Total (MT=1) and sattering (MT=2) ross-setion data of lead and bismuth, respetively (ENDF 1998).

35 21 We use thermal/epithermal neutron absorbers to ut off undesirable low energy neutrons and to prevent unwanted gamma ray prodution in the exposure ave. In this study, we designed the FNIS using a euteti alloy of 54.9% bismuth and 44.5% lead overed with boral plates to absorb any thermal neutrons. The boral TM plate is a thermal neutron absorber material, omposed of boron arbide (B 4 ) and 1100 alloy aluminum (Brooks & Perkins 1983). The boral plate is a laminar omposite of aluminum and boron arbide, onsisting of three distint layers. The outer layers of ladding are solid 1100 alloy aluminum, and the entral layer onsists of fine boron arbide partiles bound tightly within an aluminum alloy matrix. It is a good shielding material for thermal neutron beause the boron arbide provides a high neutron absorption ross-setion for thermal neutrons and is stable, strong, and durable. In this study, m-thik boral plates were used to make a boral box outside the exposure ave as shown in Fig. 2.8, as well as a boral plate plaed in front of the ell window in the pool side. For the boral box, eah side of the box was ut from a large piee of a boral plate, penetrations were made, and then the box assembled using steel frames to support the weight of the FNIS.

36 22 Fig Boral box onstruted for shielding against thermal neutrons Exposure ave To deliver optimum neutron spetra inside the exposure ave, assemblies funtioning as beam filters were onstruted behind the aluminum ell window as shown in Fig The NS reator operates with the void box installed on the front row of the grid plate, whih fills the 7.62 m-gap between ore fae and irradiation window to serve as a onethrough transmitter for fast neutrons into the irradiation ell. The detailed speifiations of the FNIS, inluding types and thikness of filters, were determined by simulating the irradiation ell, exposure ave, ombination of filters, and reator ore with MNP version 5 alulations, using ENDF/B-IV ontinuous neutron ross-setion data as desribed in Setion 2.5.

37 23 One of the design requirements of the exposure ave is to deliver neutron dose to the samples with photon dose ontamination of less than 5%. To meet with this requirement, a ube was onstruted as shown in Fig. 2.9, with outside dimensions of 58 m in height and in width and 69 m in depth, using 10-m thik briks of the euteti lead-bismuth alloy (44.5 w % Pb, 55.5 w % Bi) mentioned in Setion to ut off inident gamma rays with a minimal absorption of fast neutrons. To minimize radiation streaming through the gap of the briks, eah brik was mahined to have a 2.54 m-wide and 0.16 m-deep groove entered in the 10 m-wide fae and running its full length. During assembly, a lead strip, 2.54 m wide and 0.32 m deep, was plaed into the groove. A steel frame was installed inside the system to attenuate gamma rays streaming through the edges of the exposure ave. The inner dimension of the exposure ave is a ube 38 m 38 m 49 m that an aept small rodents, as well as a radiation monitoring system. In addition, a square penetration hole 15 m on eah side was made through the bak wall of the exposure ave for sample loading and unloading. Although the boral plate, installed in front of the irradiation ell window, absorbs thermal neutrons from the reator and pool, the neutron thermalization proess in the onrete walls of the irradiation ell produes other thermal neutrons. Therefore, some areas of the irradiation ell were lined with boral plates to ut thermal neutrons emerging from the onrete. In order to deliver neutron beams to inside of the ave with minimum gamma dose, the lead-bismuth ave was put inside a box made of boral plates supported by steel frames (see Fig. 2.8) to minimize apture gamma rays from the lead-bismuth alloy.

38 24 Fig Exposure ave for sample irradiation, whih was onstruted with the lead-bismuth briks. The penetration hole was onneted to the sample transfer system and the internal steel frame installed to support the weight of the ave. In addition, three steel bands were installed outside the lead-bismuth briks to bind all briks together. Tissue equivalent and graphite ion hambers as desribed in Setion 2.4 were installed inside the system to monitor total and photon dose rates. All of these omponents were put on a steel table that an move the exposure ave to any loation of interest based on desired dose rates. Four steel wheels installed in the steel table an sustain a weight of one ton per a wheel, so that the table an support a total weight of 4 tons. A trak was onstruted with steel H-beams to hold the steel table. Rail stops were installed at the ends of the trak for safety. Finally, the table and the trak were moved down to the irradiation ell and were installed in front of the ell window at the irradiation ell as shown in Fig

39 25 ell Window Detetors boral box Penetration Trak Fig omponents used to set up the exposure ave inside the irradiation ell, and final setup of the FNIS inluding the exposure ave, movable working table, trak, and detetors installed inside the irradiation ell.

40 SAMPLE TRANSFER AND ONTROL SYSTEM We need to provide a sophistiated ontrol system for sample transfer, whih an transfer living organisms or tissues with minimum dose reeived while they are in transit through the unfiltered neutron environment, and preventing damage whih might be done by a sudden aeleration/deeleration. To ahieve rapid transfer of samples without injury, a speially designed pneumati transfer system, apable of sending a 9 m-size sample arrier (outer diameter) into the exposure ave in a few seonds and of bringing it bak to the upper researh level, was installed as shown in Fig The sample loading box loated in the upper researh level was designed to sustain negative pressure for minimizing any leakage of airborne radioative materials. A bypass pipe was installed also to redue the traveling speed of sample arrier approahing the sample loading box from the irradiation ell. As shown in Fig. 2.12, two sensors, whih were made of miro-swith and infrared- LED/photo-transistor, were installed in the exposure ave and in the upper researh level, respetively, to ontrol the blower units, as well as the exposure time. The sample transfer pipe was installed through a small penetration between the top of the irradiation ell and the seond level of the irradiation ell. Sine there are several tortuous turns, bent or urved 45-degree radii PV pipes were used to prevent any blokage in the sample transfer pipe during the sample movement. To minimize any mehanial shok to samples or animals, round shaped sample arriers will be used. Fig shows the overall onfiguration of the sample transfer system installed in the upper ell, as well as in the lower ell.

41 27 Door for Sample Loading & Unloading Sample arrier Blower Blower For Sustaining Negative Pressure LED/Photo-transistor Sensor for Blower ontrol To Irradiation ell Valve for Speed ontrol To Irradiation ell Sample arrier To Upper Researh Level Wire Mesh Blower Exposure ave Miro-swith Sensor for Exposure ontrol Fig Diagram of the sample transfer system installed in the upper researh level (top) and in the irradiation ell (bottom) for sample loading/unloading and exposure ontrol.

42 28 Fig Home-made sensor made of miro-swith (left-hand side) and LED/phototransistor (right-hand side) for blower ontrol, as well as for exposure time ontrol. 45 degreeradii PV To Upper researh Level Fig Sample transfer system installed in the upper researh level and in the irradiation ell for sample loading/unloading and exposure ontrol. After installing the PV pipe through the penetration between the upper ell and the lower ell, 8-m thik boron-ontaining polyethylene panels were installed for minimizing neutron streaming around the penetration and the PV pipe as shown in Fig.

43 In addition, portable wooden boxes ontaining high-density polyethylene partiles were onstruted to minimize the neutron streaming through the onrete door between the upper ell and the lower ell. Fig shows the sample transfer box installed in the upper researh level for sample loading and unloading. Boron-ontaining polyethylene High density polyethylene box Fig Boron-ontaining polyethylene panel and portable wooden box for minimizing neutron streaming through the penetration and the onrete door. Relay module Loading/unloading over Fig Sample transfer box installed in the upper researh level for sample loading and unloading.

44 30 As shown in Fig. 2.16, a omputerized ontrol system was onstruted using Visual- Basi TM to manipulate eah module of the system automatially. Data olletion and blower ontrol were performed using a digital data interfae ard (PI-DIO24 TM manufatured by Measurement omputing orporation), as well as a relay module as shown in Fig The dose monitoring data were olleted using RS-232 serial ommuniation ports in the omputer. Preset values of exposure time, or neutron absorbed dose an be used to initiate sample retrieval by the pneumati system. Fig omputerized ontrol system for manipulating eah module of the FNIS system automatially.

45 31 After the sample is loaded inside the exposure ave using the sample-loading blower, the miro-swith installed inside the exposure ave an send a signal for sample exposure and then for shut the sample-loading blower off to the data interfae ard. After exposing samples for some time, the sample-unloading blower is ativated by the signal transmitted through the data interfae ard. The sample unloading blower is shut automatially when the samples pass through the LED/photo-transistor sensor loated in the upper researh level. Data interfae ard Fig Digital data interfae ard (PI-DIO24 TM manufatured by Measurement omputing orporation), whih was used for data olletion and blower ontrol.

46 RADIATION MONITORING SYSTEM Paired Ion hamber Method Radiation biology experiments require determination of both the neutron spetrum, and the absorbed dose due to neutrons and gamma rays. To assure that neutron dose to the samples is delivered with photon dose ontamination of less than 5%, we need to monitor the total and the neutron dose rates in the exposure ave during normal operation with the paired ion hamber method (IRU 1971). To aomplish this task, tissue equivalent and graphite ion hambers were onstruted as shown in Fig For total dose monitoring, a propane filled hamber made of tissue equivalent plasti (i.e., A- 150 tissue equivalent plasti, 1.27 m in diameter), with guard eletrodes around the opper olleting eletrode, was used. Graphite is usually onsidered as an air-equivalent material and has a minimal neutron absorption ross-setion. For photon dose monitoring, the wall of the seond ion hamber was made with graphite, 1.27 m in diameter, and argon gas was used as the fill gas to minimize neutron response. To determine the detetor operating voltage, the detetor readings were monitored with inreasing applied voltage while exposing the detetors at 1 Gy/min (250 kvp and 10 ma) with an x-ray mahine (Norelo MG300) at the Texas A&M Nulear Siene enter as shown in Fig The upper region of the operating voltages was hosen due to the higher ionization density along traks of seondary partiles produed by neutrons (IRU 1977). These detetors were, then, alibrated with the Piker o-60 teletherapy soure at the Texas A&M Shool of Veterinary Mediine as shown in Table 2.2.

47 33 Spaer Guard Eletrode oaxial able for signal output Srew opper Tube for Gas Filling Sealed BN Aluminum asing Graphite or Tissue Voltage Equivalent Plasti opper Eletrode Plasti Insulator O-Ring Ground opper Tube For Gas Filling BN onnetors for HV and Signal Graphitewalled Ion hamber Tissue Equivalent Plasti Ion hamber Fig Design sheme and top/side views of the tissue equivalent and graphite-walled ion hambers used in this study

48 34 x-ray mahine 1 Gy/min Dose Point Fig X-ray mahine (Norelo MG300) of the Texas A&M Nulear Siene enter used for determining operating voltage, whih set at 250 kvp and 10 ma. Data-Logging omputer Pio-Ammeter Voltage Supply Fig Keithley TM 485 Auto-ranging Pio-ammeter, Hewlett Pakard TM HP6516A D Power Supply used in this study.

49 35 The entral olleting eletrode of the detetor was onneted to a Keithley TM 485 Auto-ranging Pio-ammeter for the graphite-walled ion hamber and to a Keithley TM 6487 Auto-ranging Pio-ammeter for the tissue equivalent plasti-walled ion hamber, respetively, and the outer wall of the detetor was linked to the D power supply of the Hewlett Pakard TM HP6516A as shown in Fig Negative voltage was applied to the detetor wall to play a role as a athode. A omputer was used to display the measured dose rates based on the alibration fators and to log the measured data for future use. The detetors were inserted into aluminum asings, whih were grounded to protet the internal parts from any eletrial and physial impats. A small opper pipe as shown in Fig was installed on top of the aluminum asings to fill the detetors with their respetive gas. The gas filling system has a vauum pump and valves as shown in Fig that an be used to fill a detetor to any pressure from atmospheri to about 5 Torr. After making sure that the detetors are vauum tight, then the detetors were filled with gas of interest to 760 Torr through the 0.32 m-diameter opper tube. When filled the gas of interest, we used a speial tool to pinh off the tube to make a vauum tight seal. Fig shows the harateristis of the tissue equivalent and arbon-walled ion hambers with varying voltages. Sine the leakage urrent should be minimized to measure pio-ampere level events, a oaxial able was diretly inserted into the insulating plasti and onneted into the olleting eletrode in order to at as a able for signal output, as well as a guard eletrode. Regarding the detetion sensitivities of the detetors, the leakage urrent of eah detetor was measured with the Keithley TM 485 Auto-ranging Pio-ammeter

50 36 without any external radiation soure, and then the minimum dose rate, whih the detetor an measure, was alulated as follows (Knoll 2000); D fi ws l m = (2-1) T0 P ρ( )( ) V T P 0 Where, D is the absorbed dose rate of tissue in Gy/min, I l is the leakage urrent in pioampere, f is the onversion fator, w is the average energy loss per ion pair produed in the given gas (ev/ion pair), Sm is the relative mass stopping power of the tissue to that of the given gas, V is the detetor volume in m 3, ρ is the density of the gas at STP in kg/m 3, P 0 and T 0 are standard pressure and temperature (i.e., 760 Torr and K), and P and T are gas pressure and temperature. Both of the detetors showed the leakage urrent ( I l ) of less than 1 pa (usually, 0.3 pa), and the sensitivity level for dose rate in tissue was alulated as less than 0.2 Gy/min. Table 2.2. alibration fator for the paired-ion hambers used in this study. Detetor arbon-walled Tissue Equivalent Dose Rate for alibration (Gy/min)* Detetor Response (pa) ± ± ± ± * Dose rate in tissue measured at 0.5 m below of the water tissue. alibration Fator (Gy/min-pA) ± ± ± ±

51 37 Pressure Gauge Propane Gas Tissue Equivalent Plasti Ion hamber Fig Gas-filling system used in this study for filling argon and propane gas into graphite and tissue equivalent plasti ion hambers, respetively. 2.0 Tissue Equivalent Plasti Graphite 1.5 urrent (na) Applying Voltage (V) Fig Measured urrent with varying voltages for setting the operating voltages of the detetors used in this study.

52 38 In a mixed radiation field suh as the irradiation ell at the NSR, the paired ion hamber method desribed in the IRU-26 (IRU 1977) an be used to evaluate the separate absorbed doses of neutrons and of photons as mentioned in the previous paragraph. The tissue equivalent ion hamber, denoted by T subsript in the following equations, was used to measure responses from neutrons, as well as from photons. On the other hand, the graphite-walled ion hamber, denoted by U subsript in the following equations, was used to measure responses mainly from the photons with a lower sensitivity to neutrons. The responses of the tissue equivalent and graphite-walled ion hambers in dose rate, whih were measured with the alibration fators to the gamma rays used for the detetor alibration, an be denoted as R ' T and R ' U, respetively, and then these values an be related with the absorbed doses in tissue of neutrons and of photons as follows (IRU 1977); R = kd + hd (2-2) ' T T N T G R = k D + h D (2-3) ' U U N U G Where D N and DG are the absorbed dose rates in tissue of neutrons and of photons in the mixed radiation field, k T and ku are the ratios of the responses of the detetors to the neutrons to the responses of the detetors to the gamma rays used for the alibration, and ht and U h are the ratios of the responses of the detetors to the photons in the mixed radiation field to the responses of the detetors to the gamma rays used for the alibration. Assuming that h T and h U are lose to unity, the absorbed dose rates in tissue

53 39 of neutrons and of photons in the mixed radiation field (i.e., D N and D G ) an be obtained by D N = R k ' T t R k ' U U (2-4) D G kr = k ' ' t U U T t k R k U (2-5) Stainless Steel-Walled Ion amber Two stainless steel-walled ion hambers were, in addition, used to monitor gamma dose rates inside the irradiation ell without the FNIS. These detetors were manufatured by the LND In. for high gamma dose monitoring purpose. Sine these detetors are filled with nitrogen at 100 Torr, they are not appliable for a mixed radiation field, whih results in a high neutron dose rate ompared with the gamma dose rate, due to the apture reations of the nitrogen with neutrons suh as (n, p) and (n, α) reations. However, they an be appliable for monitoring gamma dose rates in the ase of omparable magnitude of the photon dose ompared to the neutron dose suh as the irradiation ell without the FNIS. By omparing the alulated neutron responses of two different types of filling gases (i.e., nitrogen and tissue equivalent gas) with the measured responses, we an subtrat the neutron ontribution to the reading, and obtain a better estimate of the photon dose rate.

54 Ativation Foil Method for Determining Neutron Spetra To haraterize the neutron energy spetra inside the exposure ave, foil ativation method was used with a bare aluminum-gold foil and 10 admium-overed ativation foils (Al, Fe, Au, Ni, Ti, Zn, In, o, Mg, and u) with different threshold energies that over the energy range from thermal neutrons to 13 MeV as shown in Table 2.3. In order to haraterize fast neutrons effetively, all of the foils were overed with admium plates to ut off thermal neutrons as shown in Fig Thermal neutrons were measured with bare aluminum-gold foil. After exposing the foils with the NSR against the irradiation ell with and without the FNIS, the foils were ounted with a HPGe spetrosopy system to measure speifi ativities at the end of irradiation. The flux foils ranged in thikness from mm to mm to minimize the flux perturbation and self-shielding effets of the foils (Shumaher and Randall 1975). Although all of the foils were ut to less than 1 m diameter, the spatial resolution of the flux foil sets is limited due to the larger size of the set, as shown in Fig Therefore, the flux foils were used to obtain average neutron spetra over a region of interest. The SAND (Spetrum Analysis by Neutron Detetors)-II ode (Griffin and Kelly 1994), a spetrum deonvolution ode, was used to determine a full energy spetrum from the measured speifi ativity data of the foils. A solution aeptane riterion of 10% was used to obtain neutron spetra using the measured speifi ativities of flux foils i.e., the SAND-II results were onsidered as aeptable if the standard deviation of the measured-ativity to the simulated-ativity ratio was less than 10%.

55 41 Table 2.3. harateristis of the flux foils in terms of neutron reation. Manganese Aluminum Iron obalt Nikel opper Zin Indium Gold Titanium Element Reation Half-Life Threshold Energy (MeV) 24 Mn(n,p) 24 Na 15.0 h Al(n,α) 24 Na 15.0 h Fe(n,p) 54 Mn 2.56 h Fe(n,γ) 59 Fe 44.6 d Fe(n,p) 54 Mn 2.56 h o(n,γ) 60 o 5.28 y Resonane Reation 58 Ni(n,p) 58 o 71.3 d Ni(n,p) 60 o 5.28 y u(n,γ) 64 u 12.7 h Resonane Reation 63 u(n,α) 60 o 5.28 y Zn(n,p) 64 u 12.7 h In(n,2n) 115m In 4.50 h Au(n,γ) 198 Au d Resonane Reation 46 Ti(n,p) 46 S 83.8 h Ti(n,p) 47 S h Ti(n,p) 48 S 44.0 h 4.3 admium-overed Foils Bare Al-Au Foil Fig Neutron ativation flux foils, whih onsist of admium-overed foils suh as Al, Fe, Au, Ni, Ti, Zn, In, o, Mg, and u and a bare aluminum gold foil. Eah foil paket is approximately 4 4 m.

56 MONTE ARLO MODELING Overview The MNP version 5 ode with a set of ENDF/B-VI ontinuous neutron rosssetion libraries was used to model the NSR ore against the irradiation ell at the full power operation (i.e., 1 MW). The MNP model of the NSR ore itself had already been tested by omparing the alulated results with those of other approahes, i.e., (1) an in-house three-dimensional diffusion ode (i.e., SUNMAN) and (2) foil ativation measurement (hen 1997; Kim et al. 2002). The neutron flux of the ell at 100 kw based on MNP simulation and flux foil measurements agreed well and there was no systemati differene between the MNP and flux foil results onsidering omplexity of the NSR ore and various approximations made in the MNP models (Jang et al. 2002). In addition, MNP simulations with varying ombinations of filters inluding different thikness have been performed for a feasibility study of performing boron neutron apture therapy (BNT) with the irradiation ell of the NSR, and ombinations of filters have been simulated to deliver epithermal neutron beams into the ell (Jang et al. 2003). However, the above study was foused on relatively broad epithermal neutron beams for the BNT. In this study, fast neutrons were modeled in detail with 1MeV energy intervals using the MNP F4 tally for neutron flux and F6 tally for air dose rate inside the system Reator ore Modeling This study modeled the NSR ore in as muh detail as possible, expliitly modeling

57 43 the ore as shown in Fig exept for the Sb-Be soure, whih is ignored due to its negligible effet on neutron flux. Fig shows a fuel element (left) and a fuel bundle (right) as modeled in this study. A fuel element has 38.1-m long, m diameter ative fuel setion or fuel meat. The ative fuel setion ontains a m diameter zironium rod at the enter. The fuel setion and top and bottom graphite slugs are ontained in a m thik stainless steel ladding. Top and bottom fittings of a fuel element are made of stainless steel. A fuel bundle onsists of an aluminum top handle, four rods and an aluminum bottom adaptor. The four rods in a bundle an be four fuel elements, three fuel elements + an instrumented fuel element, three fuel elements + a ontrol rod, or two fuel elements + two water elements as shown in Fig The top handle onsists of a handle, a loking plate and a loking bolt. The bottom adaptor has a shape to fit into the grid plate and the top of the adaptor ontains four tapped holes into whih fuel rods are threaded. The details of these ompliated top handle and bottom adaptor were not expliitly modeled in this study. The NSR ore utilizes FLIP (Fuel Life Improvement Program) type fuel, in whih zironium hydride moderator (ZrH 1.65 ) is homogeneously ombined with 8.5 weight % of uranium (70% enrihed in 235 U) and approximately 1.5 weight % of erbium (as burnable poison) for new fuel. urrently, the fuel bundles in the NSR ore ontain 67-84% of their original ontent of 235 U. In this study, the fuel bundles in the ore were divided into three groups aording to their 235 U depletion levels. Then, all the fuel elements in the same group were assumed to have the same amount of 235 U. The vertial variation of fuel depletion was modeled based on the vertial variation of thermal

58 44 neutron flux in the NSR ore. To inorporate the vertial variation of fuel burn-up, the fuel meat setion of a fuel element was divided into five segments as shown in Fig and the onentrations of 235 U and 167 Er were determined for eah of these segments. In this study, only 167 Er onentration was determined for the depletion of burnable poison sine 166 Er onentration does not hange muh over the lifetime of the fuel due to its small neutron absorption ross-setion. In this study, we did not expliitly model the fission produt in the NSR ore due to the lak of a detailed inventory. A rough approximation was made. First, the fission produts in the NSR ore were represented by so-alled average fission produt (Briesmeister 2003; Foster and Arthur 1982). The "average fission produt" is a pseudofission produt that has a lump fission produt ross-setion representing the rosssetions of the fission produts that are produed from one 235 U fission. The onentration of the average fission produt in the reator ore was then determined by repeating ritiality alulations; that is, the onentration of the average fission produt was adjusted to make the reator ritial for a given ritial ondition of the irradiation ell run (i.e., shim safety rods at 80%, transient rod at 100%, and regulating rod at 50%).

59 Fig MNP drawings denoting NSR, void box, ell window, irradiation ell, and the exposure ave on x-y plane, y-z plane, and x-z plane. 45

60 46 Top fitting (stainless steel) Graphite Top handle (aluminum) Zironium rod Fuel meat Setion (38.1 m) Fuel meat (diameter = m) ladding (0.051m thik stainless steel) Top view of a fuel bundle Graphite Bottom fitting (stainless steel) Bottom adapter (aluminum + water mixture) Fig Fuel element and fuel bundle as modeled in this study. Note that the fuel meat setion is divided into 5 segments and this study alulates neutron flux in eah of these five segments (Kim et al. 2002).

61 47 Shim safety rod Transient rod 12.7 m (void) Regulating rod m (poison) m (poison) m (poison) m (fuel) m (void) m (void) Fig Monte arlo modeling of ontrol rods as modeled in this study (Kim et al. 2002).

62 48 The reator power is ontrolled by a total of six ontrol rods (4 shim safety rods, a regulating rod, and a transient rod). Fig shows the ontrol rods as modeled in this study. A shim safety rod, whih has sram apability, onsists of a poison setion, a fuel setion and two void setions on the top and bottom. The poison in the shim safety rod is borated graphite (B 4 25% and graphite 75% by weight). The diameter of the poison is m and the thikness of the stainless steel (SS304) lad is m. The fuel setion is almost idential to a regular fuel rod exept that the fuel meat is m in diameter. The transient rod, whih is used for reator pulsing and has sram apability, onsists of a poison setion and a void setion. The poison used in the transient rod is also borated graphite. The diameter of the poison is m and the thikness of the stainless steel (SS304) lad is m. The regulating rod, whih is used for 'servo' ontrol of reator power, is omposed of a poison setion. The rod size is the same as the transient rod. The only differene is that the poison in the regulating rod is pure B 4 powder. These shim safety rods, transient rod and regulating rod were raised to 80%, 100% and 50%, respetively, to model the reator against the irradiation ell at the ritial ondition of 1-MW operation. This study did not model the Sb-Be soure, whih is loated at 8 in Fig due to its negligible effet on neutron flux. The power monitoring detetors and pneumati devies were modeled as void spae. The grid plate, whih is made of aluminum with 54 holes, was modeled as a homogenous mixture of aluminum (84 weight %) and water (16 weight %). The reator ore was modeled as surrounded by water gap in the opposite side of the irradiation ell, whih is big enough to aount for full neutron sattering in

63 49 water. To model fuel meat setions, this study used the ENDF/B-VI neutron ross-setion data whih were evaluated at high temperatures (558 K or 600 K). These hightemperature ross-setion data were obtained from the Korean Atomi Energy Researh Institute (KAERI). The average temperature in the fuel elements (when the reator is operating at 1 MW) is approximately 550 K and, therefore, the neutron ross-setion data were adjusted for the temperature differenes using TMP ards in MNP. For other part of the reator ore, this study used the standard ENDF/B-VI ross-setion data that were evaluated at 300 K. The slow neutron sattering law S(α, β), whih was evaluated at 300 K or 600 K as appropriate, was used to aount for the moleular binding effets in ZrH, water and graphite Irradiation ell Modeling Inluding the FNIS In the pool side, a stainless steel (Type 304) liner is installed in the reator pool for water ontainment and water purity. However, there is no steel liner inside the irradiation ell. The upper 5.2 m of the wall is made with the standard onrete, and the lower portion of the wall is barites onrete and light onrete (NS 2003). Therefore, the wall of the irradiation ell has a good apability for reator experiments in terms of neutron and gamma ray shielding with the maximum reator power of 1 MW. There is, however, a penetration of 50 m by 30 m size for ooling water piping and eletrial ables, and therefore additional shielding around the penetration is neessary for further dose redution in the upper researh level. The penetration on top of the ell will be used

64 50 for sample loading and unloading as desribed in the Setion 2.3. This study did not model the penetration of top of the ell. Although there are several steel H-beams in the ell for strutural integrity and there are unused ooling omponents, it was not modeled in this study. Fig shows the MNP drawings denoting the exposure ave, movable steel table, extra lead briks, aluminum plates, and ell window. a b Fig MNP drawings denoting exposure ave surrounded by a boral box, a movable steel table, a ell window, and the walls of the irradiation ell on a) x-y plane, b) y-z plane, and ) x-z plane.