Abstrat The goal of this qualifying projet was to investigate the output of neutron radiation by the new fast neutron failities at the University of M

Transcription

1

2 Abstrat The goal of this qualifying projet was to investigate the output of neutron radiation by the new fast neutron failities at the University of Massahusetts, Lowell Researh Reator. This was done by using Monte Carlo-based radiation transport simulation software to prepare a set of dosimetri studies that were later arried out at the Lowell fast neutron failities. A linear relationship between reator power and partile output was found.

3 Aknowledgements I would like to thank Professor Blake Currier for his guidane and support throughout this projet and shool year, even as he went through his first months as a father. Thank you to Professor Dave Medih for weloming me into his lab and offering lasses that proved invaluable while working on this projet. Thank you to Professor Germano Iannaione for introduing me to the folks at the WPI radiation lab. Thanks also to Steve Snay, Leo Bobek, and the staff of the University of Massahusetts, Lowell Researh Reator for granting aess to the reator failities and giving me a great hane to do some really ool experiments. A big thank you to my parents for their support throughout this and every year. Finally, shout outs to boots and also to ats. All-nighters are easier with good musi.

4 Exeutive Summary Nulear siene is a popular area of study with many appliations in the fields of researh and energy: 104 nulear power plants and 54 researh reators were ative in the United States alone as of 2011 [1, 19]. The reator at the University of Massahusetts, Lowell is one suh researh reator. In order to perform experiments using its reently installed fast neutron faility, it is essential to understand the output of partiles from the reator at various power settings. The goal of this major qualifying projet (MQP) was to investigate the output of neutron radiation by the University of Massahusetts, Lowell Researh Reator (UMLRR), and to independently evaluate the safety of the reator failities. Monte Carlo-based omputational models were used to simulate different physial geometries and the behavior of neutron radiation therein. The data obtained from the simulations was used to plan a study of the fast neutron failities at the UMLRR. In this study, dosimeters were exposed to radiation at the fast neutron faility and measured the photon and neutron dose. The reorded dose is diretly related to partile fluene. The dosimeters used aluminum oxide thermoluminesent detetion material for photon detetion, and CR-39 polyethylene material, whih reords ion traks aused by interations with neutrons of energies between 40 kev and 40 MeV, to measure neutron dose [11]. Dose measurements were taken for four different reator power settings for varying lengths of time to establish a relationship between reator power and reorded dose rate. A linear orrelation was found between the power setting of the reator and the ombined gamma and neutron dose rate reorded by the dosimeters. It an be onluded that the fluene rate of partiles inreases linearly with reator power, though more data is required to establish an exat relationship.

5 Table of Contents Abstrat... 1 Aknowledgements... 2 Exeutive Summary... 3 Table of Contents... 4 List of Figures... 6 List of Tables... 8 Introdution Bakground Radiation and Dosimetry Nulear Transformations Ativity Radiation Quantifiation Radiation Interation with Matter Thermoluminesent Dosimeters Neutron Dosimetry Radiation Weighting Fators and Dose Limits Biologial Effets of Radiation Neutron Soures and Energy Fission Fast Neutrons Epithermal Neutrons Thermal Neutrons Neutron Interation Cross Setion Neutron Ativation Neutron Soure Charaterization The Neutron Ativation Equation Neutron Ativation Analysis Computational Simulation of Radiation The Monte Carlo Method Methods Monte Carlo (MCNP) Analysis... 31

6 i. Determination of Number of Histories ii. Comparison of Geometries iii. Effet of Water Baking on Reorded Neutron and Gamma Dose from Thermal and Fast Neutrons Fast Neutron Dose Results Monte Carlo Results i. Determination of Ideal Number of Histories ii. Comparison of Geometries iii. Effet of Water Baking on Reorded Neutron and Gamma Dose from Thermal and Fast Neutrons Fast Neutron Dose Results Disussion Works Cited Appendix A: Effet of Water Baking MCNP Simulation Complete Result Tables Appendix B: MCNP Deks Spherial Geometry Linear Geometry Water Baking pstudy Dek... 69

7 List of Figures Figure 1 - In photoeletri absorption, an atomi eletron fully absorbs the energy of an inident photon Figure 2 - In Compton sattering, the inident photon is not fully absorbed by the atomi eletron. The angle between the sattering eletron and photon is related to the photon s initial energy Figure 3 - In pair prodution, a high-energy photon under the influene of an atomi nuleus is onverted into an eletron-positron pair. The resultant partiles propagate in exatly opposite diretions Figure 4 - The exess relative risk of solid aners for Japanese atomi bomb survivors as alulated by the BEIR VII ommittee [7]. A linear, no-threshold fit and a linear-quadrati fit are plotted Figure 5 - The deay sheme of gold-198 to stable merury-198 by beta emission Figure 6 - A possible series of interations aused by an inident neutron and the order in whih they are traked [26] Figure 7 - The spherial geometry used in parts i. and ii. The thinner layers are the tally ells.. 31 Figure 8 (A) The spherial geometry, with a entral point neutron soure. (B) The linear geometry, with dis neutron soure (line to the left of the squares) Figure 9 - A tally ell (green) with a 4m water baking (blue) within the retangular prism world ell (gray shade) Figure 10 - A Luxel+ dosimeter, with pen for sale Figure 11 - (Left) An overhead view of the neutron failities. The thermal olumn and fast neutron bunker are both visible. (Right) Another external view of the fast neutron bunker. Highdensity onrete and other insulating materials are used to ontain the radiation released within Figure 12 - A piture of the setup inside the bunker. The movable rak system (holding the solid water, right of piture) ould be moved from outside the bunker to the front of the ollimation tube (left of piture). The tube has a 30m diameter Figure 13 - The solid water positioned in front of the ollimation tube. The bag ontaining the dosimeters is held to the solid water bloks by tape Figure 14 - A plot of reported error at various tally distanes by number of histories taken. As the number of histories inreases, error dereases. No error is reported at any distane when 1x10 7 histories were simulated Figure 15 - A hundred-thousand partile sampling of energy depositions in the linear (A) and spherial (B) geometries Figure 16 - Plot of the dose reorded by the tallies in the spherial and linear geometry setups. The spherial setup had 0% reorded error; the linear setup had 0.34% error Figure 17 - Plot of partile fluene by distane from soure and water baking thikness for thermal neutron simulations Figure 18 - Plot of partile fluene by distane from soure and water baking thikness for fast neutron simulations

8 Figure 19 - Plot of effetive dose rate as a funtion of reator power, showing a linear relationship

9 List of Tables Table 1 - Radiation weighting fator by partile type and energy [25]. The values for neutron radiation are from ICRP Publiation 60 [20] and are aurate for approximation Table 2 Annual dose limits set by the ICRP and US NRC [9, 13] Table 3 - Biologial effet on humans of radiation dose [16]. Marked symptoms an our after loalized radiation exposure, not just a whole-body dose Table 4 Reator power and dosimeter exposure times for the fast neutron irradiation experiment Table 5 - Tally error values by number of histories run for five different tally distanes Table 6 - Dose and tally errors reorded for the spherial and linear setups partile histories were run for eah geometry Table 7 Seleted data from the results for thermal ( ev) neutrons impinging upon a dose tally with a water baking of various thiknesses Table 8 - Seleted data from the results for fast (1 MeV) neutrons impinging upon a dose tally with a water baking of various thiknesses Table 9 - Results of dosimetri readings taken at the University of Massahusetts, Lowell Fast Neutron Irradiation faility Table 10 - Combined (neutron and gamma) effetive dose rates for eah dosimeter and averages by reator power level Table 11 - Photon, neutron, and ombined dose readings onverted to SI units Table 12 - Exposure time required to reah signifiant dose and effetive dose. Note the varying units of time Table 13 - Results for thermal neutrons impinging upon a dose tally with no water baking Table 14 - Results for thermal neutrons impinging upon a water baking of thikness 0.1 entimeters Table 15 - Results for thermal neutrons impinging upon a water baking of thikness 1 entimeter Table 16 - Results for thermal neutrons impinging upon a water baking of thikness 4 entimeters Table 17 Results for thermal neutrons impinging upon a water baking of thikness 10 entimeters Table 18 - Results for fast neutrons impinging upon a dose tally with no water baking Table 19 Results for fast neutrons impinging upon a water baking of thikness 0.1 entimeters Table 20 - Results for fast neutrons impinging upon a water baking of thikness 1 entimeter. 63 Table 21 - Results for fast neutrons impinging upon a water baking of thikness 4 entimeters Table 22 - Results for fast neutrons impinging upon a water baking of thikness 10 entimeters Table 23 - Results for fast neutrons impinging upon a water baking of thikness 20 entimeters

10 Table 24 - Results for fast neutrons impinging upon a water baking of thikness 50 entimeters

11 Introdution With 104 nulear energy plants and 54 researh reators present in the United States alone as of 2011 [1, 19], nulear energy and experimentation is never far from the publi spotlight. From military to mediine, nulear siene is an ever-present faet of modern researh and tehnology. Researh reators are used for a variety of purposes, many involving neutron generation [19]. The neutrons produed by the nulear proesses in the reator ore an be moderated to a desired energy and then direted for various uses. One suh use is neutron imaging, where a neutron beam is used like an x-ray to view the internal struture of an objet, is a ommon proedure at many researh reators. Neutron imaging an be used to evaluate the strutural integrity of materials under a variety of onditions down to the atomi level and with greater larity than other methods, making it an exiting field of researh. Another ommon use of researh reators is the reation of radioative materials for use in industry and mediine through neutron ativation, where stable elements are exposed to and absorb low-energy thermal neutrons. One example in the field of mediine is irradiation of ytterbium-176. This proess yields ytterbium-177, whih in turn deays rapidly to lutetium-177, a radioisotope often used for radiation therapy in aner treatments. Neutron ativation an also be used to identify the omponent elements of unknown materials, as ativated materials will emit distint radiations based on their omposition. The University of Massahusetts Lowell Researh Reator (UMLRR) is a pool-type researh reator, with its ore submerged in a 35-foot deep water pool for ooling and shielding purposes. The reator is apable of produing one megawatt of power at peak output. Neutrons from the reator an be moderated to lower energies, on the sale of 0.01 ev, for use in neutron ativation and neutron imaging experiments. This takes plae in the thermal olumn, where moderating materials slow the neutrons through ollisions with atomi nulei. Only reently has the reator been outfitted with omponents that make experiments with fast neutrons with energies on the sale of 1 MeV. The plate-type low-enrihed uranium fuel used in the reator allows for high neutron fluene rates of over neutrons per square entimeter per seond [24]. It is important to gather as muh data as possible on the output of the

12 reator s fast neutron failities in order to properly evaluate its apabilities and the risk it may pose to operators. The reator failities an only be made safer with more information. The main goal of this projet was to haraterize the thermal olumn and fast neutron failities at the University of Massahusetts, Lowell using a ombination of omputational simulations of radiation transport, dosimetri readings taken at the reator, and ativation experiments performed on-site. Charaterization of the neutron failities began with simplified models made using Monte Carlo Neutral Partile Transport Code (MCNP). These models were used to study the relationship between neutron energy and absorbed dose of neutron and gamma radiation. The models were also used to investigate the effiay of introduing a solid water baking to dosimeters in order to more aurately simulate dose that would be reeived by a human. An experiment performed at the fast neutron failities onsisted of exposing neutron- and photon-sensitive dosimeters to the radiation produed by the reator at various energies and for varying lengths of time. The dosimeters ontained aluminum oxide thermoluminesent material for reording photon dose and CR-39 polyethylene for reording neutron dose. This experiment resulted in dose and dose rate readings whih were ompared to the MCNP simulations to better understand the apabilities of the reator. The data gathered with this series of experimentation on the neutron failities at UMass Lowell will be highly useful in evaluating their apabilities and safety. When working with potentially hazardous radioative material in a high-powered environment like a reator, espeially one with relatively new failities, it is good to have a muh information about your system as possible. Furthermore, the proedures performed ould easily be reprodued at any omparable reator to produe ationable data and help make nulear failities around the world safer.

13 Bakground Radiation and Dosimetry Radioativity arises in nulei that are energetially unstable, due to an irregular neutronproton ratio or beause the nuleons are not in a ground state. For example, arbon-12 is a nuleus with 6 protons and 6 neutrons and is stable. A possible isotope of arbon with two additional neutrons, arbon-14, is not stable and will transform into the stable isotope nitrogen- 14 by emitting a beta partile. Radioative isotopes are also referred to as radionulides. Nulear Transformations Different transformations an our depending on how unstable a nuleus is. A nuleus with a neutron-to-proton ratio that is too high to be stable an emit a β - partile, an eletron, to redue the ratio and emit energy in the form of the eletron's kineti energy. When a nuleus has a low neutron-to-proton ratio and needs to release a high amount of energy, it will emit an alpha partile, omprised of two protons and two neutrons. If it needs to release a lower amount of energy, it an emit a β + partile, a positron. If the nuleus is not energeti enough for even that, it an undergo eletron apture, where an orbital eletron is absorbed into the nuleus to redue the neutron-proton ratio, and then emit a photon to shed energy. Many, if not all, nulear transformations result in the nuleons having a more energeti onfiguration than the ground state. An exited nuleus will naturally shed energy and transition to a lower energy state by either emitting a photon, or by imparting energy to an orbital eletron and thereby ejeting it. In eah of these transformations, the emitted partiles have enough energy that they ould ejet eletrons, protons, or neutrons from atoms that they interat with. As suh, alpha, beta, and gamma are known as ionizing radiations. The effets these radiations an have on other atoms an break hemial bonds or ause the formation of highly reative ions. In materials this an ause a breakdown of the atomi struture; in living tissue, it an lead to ell death or the reation of anerous ells.

14 Ativity The number of transformations a group of radionulides undergoes per unit time is known as its ativity. The SI unit for ativity is the Bequerel (Bq), equivalent to one transformation per seond. Another ommon unit of ativity is the urie (Ci), equivalent to 3.7x10 10 Bq. The ativity of a given radionulide is proportional to the amount of the nulide present: A = λn (1) As a radionulide transforms into more stable nulides over time, the number of atoms of the original nulide dereases exponentially. This means that the ativity of the sample due to that nulide also dereases: N = N 0 e λt (2) A = λn 0 e λt (3) The proportionality onstant λ is known as the deay onstant, and is speifi to eah nulide. Radioative transformations on the atomi level are entirely stohasti; that is, it is impossible to say when a given unstable atom will transform. On a marosopi level however, radioative elements have a half-life, the amount of time it takes for half of the atoms of a radionulide to transform. Half-life is inversely proportional to the deay onstant, as T1/2 = ln 2 λ. Radiation Quantifiation Dosimetry refers to the measurement of radiation dose reeived by material or tissue due to exposure to ionizing radiation. Radiation is seen by many as a highly dangerous toxiant, responsible for radiation poisoning, aner, and a number of other less-than-desirable maladies. The truth is that, while radioative material an be dangerous if mishandled or misused, radiation is quite simple to detet. Its effets are well-known and very small doses an be measured. There are a variety of ways to quantify radiation levels. The first way radiation was ever quantified, a value known as exposure, is the measurement of ionization aused by photons in

15 air, ommonly measured in roentgen (R), equal to 2.58x10-4 oulombs per kilogram of air, or 1 eletrostati unit per ubi entimeter. This type of measurement was designed to measure photon radiation in air speifially. Many types of radiation detetors are gas ionization hambers that use this type of measurement to give the user an idea of the radiation ativity and energy in an area. A more ubiquitous measurement is absorbed dose, whih quantifies the amount of energy imparted to a material by ionizing radiation. The unit of absorbed dose is the gray (Gy), defined as one joule per kilogram. Sine absorbed dose is measured in energy per unit mass, any given point in an objet not neessarily the objet as a whole an be assigned an absorbed dose. Radiation Interation with Matter Radiation interats with matter in a variety of ways depending on what type of radiation it is. In partiular, unharged partiles photons and neutrons interat very differently from harged partiles suh as beta and alpha partiles. Charged partiles interat through Coulombi fores, leading to what is known as diret ionization. Unharged partiles ause indiret ionization through nulear interation. In general, unharged partiles interat with materials far less than harged partiles. This means that unharged partiles penetrate further into any material they enounter [22]. Photons may be produed either in an atomi nuleus or in the eletron loud surrounding it. When an exited and unstable nuleus ollapses to a more stable onfiguration, a photon will be emitted as the arrier of the energy differene. Photons emitted from an atomi nuleus are known as gamma rays. Photons may also be emitted by eletrons in an atom whih move to a lower energy state; the differene in energy is released as a photon. Regardless of their origin, these photons an have a wide range of energies. The distintion between gamma rays and nongamma photons is made beause gamma photons are reated with harateristi energies that an provide information on the atom from whih it ame. Photon interation with matter has three main modes: photoeletri absorption, Compton sattering, and pair prodution [22]. Photoeletri absorption ours when a photon is entirely absorbed by an atomi eletron. With the energy from the photon, the eletron will be raised to a

16 higher-energy orbital or may be ejeted from the atom entirely (Figure 1). An ejeted eletron will go on to have further interations with the material and is an example of seondary radiation. This photoeletri effet is most ommon with photon energies lower than 0.1 MeV. Figure 1 - In photoeletri absorption, an atomi eletron fully absorbs the energy of an inident photon. Compton sattering is the dominant effet with photons with energies between 0.1 MeV and 1 MeV. Compton sattering ours when an inident photon interats with an atomi eletron and imparts a portion of its energy suffiient to ejet the eletron from the atom. The result is a seondary eletron and a lower-energy photon whih propagate at an angle to one another that may be related bak to the initial energy of the photon. Both the lower-energy photon and the seondary eletron will go on to have further interations (Figure 2). Figure 2 - In Compton sattering, the inident photon is not fully absorbed by the atomi eletron. The angle between the sattering eletron and photon is related to the photon s initial energy.

17 When a very high energy photon omes under the influene of the eletromagneti field of an atomi nuleus, it may be onverted into an eletron-positron pair. Eletrons and positrons both have a rest mass of 511 ev; thus the original photon must have an energy of at least MeV, the sum of the rest energies of the resulting partiles. All energy in exess of the required MeV is divided equally between the pair as kineti energy, and the partiles will propagate in exatly opposite diretions (Figure 3). Both partiles will interat with surrounding matter through eletromagneti fores until they shed suffiient energy. The eletron may be aptured by an atom, whereas the positron will enounter another low-energy eletron and be annihilated. Figure 3 - In pair prodution, a high-energy photon under the influene of an atomi nuleus is onverted into an eletron-positron pair. The resultant partiles propagate in exatly opposite diretions. Neutrons interat with matter by olliding diretly with the nuleus. When a neutron ollides with an atomi nuleus, it will satter elastially and possibly impart enough energy to knok free a neutron or proton. If a neutron is knoked free, it will ontinue on muh like the original inident partile, albeit with a lower energy. The reoiling atom an also ollide with surrounding partiles. One a free neutron reahes a low enough energy, it may be aptured by a nuleus, possibly resulting in a nulear reation. A freed proton will interat as a harged partile with atomi eletrons. Through Coulombi fores the proton an impart some of its energy to an eletron, whih may be enough to free it and ionize the atom. Even if the eletron is not freed, it will be greatly exited and emit a high-energy photon when it falls bak to a stable energy level.

18 Thermoluminesent Dosimeters Thermoluminesent dosimeters (TLDs) are radiation dosimeters whih measure dose using rystal strutures, ommonly aluim fluoride or lithium fluoride, with intentional impurities (alled a dopant, or doping material) to apture exited eletrons in exited energy states. When the TLD is exposed to ionizing radiation, eletrons are exited to higher energy states. Normally they would fall bak to ground state quikly, but the impurities in the rystal struture trap the eletrons in the exited state [21]. To measure the dose reorded by the TLD, the dosimeter is heated, whih relaxes the rystal struture and allows the trapped eletrons to fall bak to ground state by emitting photons. The photons emitted are measured and related to the radiation dose reeived. TLDs are mainly used to measure beta and gamma radiation. Various metal foils an be used to filter the type of radiation that reahes the dosimeter. The material used in the TLD affets what type of radiation it will reord. For example, a lithium fluoride detetor an detet gamma and neutron dose sine neutrons will interat with the lithium and produe an alpha partile that will interat with eletrons in the rystal lattie. A alium fluoride TLD will not detet neutron radiation sine neutrons do not readily interat with heavy alium atoms and are similarly unlikely to interat with eletrons diretly sine neutrons are unharged. The thermoluminesent material in the dosimeters used for the experiments at the Lowell reator was aluminum oxide that was formulated to ontain arbon impurities (Al2O3:C). Neutron Dosimetry An effetive way to measure neutron radiation dose is to use allyl diglyol arbonate, ommonly known as CR-39. CR-39 is a polyethylene resin that has been used to make lenses and eye protetion. When pure CR-39 is exposed to neutron radiation, neutrons an interat with the hydrogen atoms in the resin to ause protons to reoil with great energy [5]. These ause ion traks to be ethed into the material. In dosimeters ontaining boron, a reation whih releases an energeti alpha partile also leads to ion traks. Just as with TLDs, metal filters an be used to

19 blok neutrons of ertain energies. For example, admium is a strong absorber of low-energy neutrons, so it is used when fast-neutron dosimetry is desired. To measure the dose, the CR-39 material is ethed, ommonly with sodium hydroxide, to enlarge the traks. The traks are then optially analyzed to evaluate the dose. Where TLDs are insensitive to neutron radiation, CR-39 does not detet gamma or beta radiation. A ombination of a TLD hip and CR-39 material are used to make dosimeters whih are useful for measuring a variety of radiation. The dosimeters used in the experiments at the Lowell reator ontained CR-39 material for reording neutron dose. The CR-39 deteted neutrons with energies between 40 kev and 40 MeV. Radiation Weighting Fators and Dose Limits Sine radiations differ in both physial makeup and interation mehanism, some types are more damaging than others. An alpha partile has twie the harge and almost 2000 times the harge of a beta partile, so it is to be expeted that an alpha partile will interat more and with greater effet than a beta. One metri used to quantify the differene between radiations is linear energy transfer (LET), whih is the amount of energy transferred by a partile s interations per unit length of its trak. LET is ommonly measured in units of kev/µm [6]. In general, LET is proportional to partile mass and harge, and inversely proportional to partile energy: high energy partiles will have fewer interations per unit trak length than slower partiles. Another onsequene of the differenes in energy deposition by various partiles is a differene in how muh impat they will have on a biologial system. This differene is quantified using relative biologial effetiveness (RBE). RBE is alulated by finding the dose D250 of 250 kvp photons (photons reated with a soure of 250 kv potential) required to ause a given biologial effet, and the dose Dr required of a test radiation to ause the same effet. The relative biologial effetiveness of the test radiation is the ratio of these two values [6]:

20 RBE = D 250 D r (4) RBE and LET are not diretly related beause of the nature of the ritial target in biologial ells. Relative biologial effetiveness inreases with linear energy transfer up to an LET of around 100 kev/µm. At this point, the average distane between ionization events is equal to the width of the DNA double helix, and the radiation is most likely to ause a double strand break in the DNA, whih is the main ause of most biologial effets in living organisms. Above 100 kev/µm, RBE falls off rapidly. This is beause there are more ionization events than required to suffiiently damage the DNA, meaning energy is wasted. There are many omplex differenes in relative biologial effetiveness of different types and energies of radiation, so it is neessary to use a more generalized quantifiation system. A radiation weighting fators (wr) is now ommonly used, based on RBE studies as well as a variety of other onsideration. In these systems, low LET radiation suh as photons and eletrons have weighting fators of one. Alpha partiles have the highest weighting fator at 20. Highenergy proton radiation has a weighting fator of 2. Neutron radiation has a highly variable biologial effetiveness based on partile energy, so a ontinuous funtion is used to relate energy with weighting fator [17]. Disrete values an be used when preision is not required. Table 1 - Radiation weighting fator by partile type and energy [25]. The values for neutron radiation are from ICRP Publiation 60 [20] and are aurate for approximation. Radiation Type Partile Energy Weighting Fator w R Photon All 1 Beta All 1 (eletron, positron) Proton >2 MeV 2 Neutron <10 kev 5 10 kev 100 kev kev 2 MeV 20 2 MeV 20 MeV 10 >20 MeV 5 Alpha, Heavy Ion All 20

21 These radiation weighting fators are used as dose multipliers to alulate an effetive dose E. While the unit of absorbed dose is the gray (Gy, equal to 1 J/kg), the units of effetive dose are the Sievert (Sv). To obtain an effetive dose value, the absorbed dose is multiplied by the weighting fator of the radiation involved: E = w R D (5) Thus an absorbed dose of 1 Gy of photon radiation is an effetive dose of 1 Sv. However, 0.5 Gy of proton absorbed dose is an effetive dose of 1 Sv beause the proton weighting fator is 2. In order to protet workers in the radiation industry engineers, laborers, medial radiologists, and anyone else who regularly omes into ontat with radioative materials as well as the general publi, regulatory ommittees have set dose limits for both workers and the publi. The International Commission on Radiologial Protetion (ICRP) sets the oupational dose limit for whole-body dose at 0.02 Sv per year, averaged over 5-year intervals. The reommended limit for the publi is Sv per year. Type of Limit Table 2 Annual dose limits set by the ICRP and US NRC [9, 13]. International Commission on Radiologial Protetion US Nulear Regulatory Committee Oupational Publi Oupational Publi Effetive Dose 0.02 Sv* Sv 0.05 Sv Sv Eye Lens 0.15 Sv Sv 0.15 Sv -- Skin 0.50 Sv 0.05 Sv 0.50 Sv -- Extremities 0.50 Sv Sv -- *To be averaged over a five-year period, with provision that dose should not exeed 0.05 Sv in any one year. Limits for dose to the eye lens and skin are speified beause those tissues are not neessarily proteted by the limit put on whole body dose. Although the standing theory is that any radiation exposure will inrease aner risk, these limits have been determined to provide suffiient protetion.

22 Biologial Effets of Radiation Ionizing radiation interats with living ells primarily by damaging ell DNA. While radiation an interat with and damage other parts of a ell, it has been shown that relatively high doses are required for a biologial effet to arise due to this type of damage (on the sale of gray); whereas irradiation of the ell nuleus leads to biologial effets at muh lower doses, on the sale of entigray [6]. While a single hit by radiation of suffiient energy an ause a strand break in the DNA hain, many interations are required to ause suffiient damage to ell organelles to lead to ell death. Note, however, that thanks to the redundant struture of many hromosomes (human hromosomes ontain two opies of geneti information, for example), more than one strand break is required to ause biologial effets. Suffiient damage to the DNA hain in the nuleus of a ell an lead to aberrations in the struture of the hain, whih, if not properly repaired by ell mehanisms, an lead to abnormal ell dupliation or lonogeni death, where the ell is unable to divide properly. In the former ase, abnormal ell growth an lead to aner. In the latter, aute doses of radiation an ause ell death. A loalized dose will ause topial effets suh as hair loss or skin reddening, while whole-body dose will lead to the symptoms of aute radiation sikness. Some types of ells are more radiosensitive than others. Radiosensitivity is dependent on a number of fators, inluding ell ativity and division rate, ell age, and ell speialization. Aute radiation sikness is presented aording to these sensitivity fators. At relatively low doses, sikness presents itself over a longer time sale beause only young ells are affeted: symptoms will not arise until the urrent population of ells dies out naturally. At higher doses, more ells are affeted and symptoms are more prompt and severe. The onept of a 50% lethal dose (LD50) is used in the ontext of high-dose aute irradiation. The LD50/60 is the dose at whih 50% of individuals exposed will survive the 60 days following irradiation. If an individual survives past 60 days, the body will have mostly reovered from the ell death aused by radiation and is expeted to survive. The organs that are responsible for the reation of blood are the most radiosensitive [6]. Stem ells purposed for replenishing blood ells are ative, young, and unspeialized, and are thus the first to die when exposed to radiation. Doses as low as 0.15 gray an affet blood ount, and doses around 0.5 gray will affet white blood ell levels within a number of minutes.

23 Organs in the digestive system are the next most radiosensitive life-ritial organs [6]. Again, it is the stem ells that will replae the linings of the stomah and intestines that are most severely affeted. Doses of 5 gray or more will ause degeneration of the digestive organ ells within hours. The biologial effets of aute radiation exposure area as follows: Table 3 - Biologial effet on humans of radiation dose [16]. Marked symptoms an our after loalized radiation exposure, not just a whole-body dose. Dose (Gy) Effet Possible blood ount hanges 50 Certain blood ount hanges 100 Vomiting 150 Death threshold LD 50/60 with no medial are *Epilation (hair loss) *Erythema (skin reddening) LD 50/60 with medial are Below a 1.5 gray dose, symptoms are similar to a viral sikness: fatigue, nausea, and vomiting. The ell death aused by radiation is reognized by the body as a biologial infetion and it responds as suh. Above 1.5 gray, a dosed individual will have low ounts of red blood ells, white blood ells, and platelets. Infetion is the most serious risk at this point, and hemorrhaging is another major onern. Isolation, antibiotis, and blood transfusions will inrease likelihood of survival. Doses high enough to effetively wipe out blood prodution ells will require a bone marrow transplant for survival to be possible. Radiation doses above 5 gray may wipe out the gastrointestinal tissues. At this point, the body an no longer absorb nutrients or protet itself from bateria in the intestines, and death is ertain. A full-body dose above 20 gray will affet even the most radioresistant tissues in the musular and nervous systems. Rapid death is all but ertain, often due to edema. Interestingly, muh higher doses are required to ause death if dose is limited to the head region alone. The reason for this is not well understood [6].

24 While lower doses will not lead to non-stohasti effets suh as the aute symptoms desribed above, it is thought that any dose is liable to inrease an individual s risk of developing aner [7]. Unfortunately, this stohasti effet is diffiult to quantify. Muh of the data regarding radiation-indued aner is from the survivors of the atomi bomb strikes on the Japanese ities of Nagasaki and Hiroshima, and the Chernobyl reator meltdown. The A-bomb survivors were exposed at high dose rates to gamma radiation and a smaller omponent of neutron radiation. Chernobyl survivors, mainly those who were subjet to fallout from the disaster, were exposed to a variety of radioative isotopes, with the beta and gamma-emitting iodine-131, strontium-90, and esium-137 being the most prevalent [4]. Nulear regulatory ommittees have adopted risk models based on a linear, no-threshold relationship between radiation dose and risk of aner [7, 16]. That is, any radiation exposure at all will inrease risk of aner above the baseline, and aner risk inreases linearly with dose; an individual who reeives a 0.3 Sievert effetive dose in a given time period will be ten times more likely to develop aner than one who reeives 0.03 Sievert in the same time period. The linear, no-threshold model was found to fit existing data well, and is also simple to use and makes onservative risk estimates, enouraging minimal radiation exposure. Figure 4 - The exess relative risk of solid aners for Japanese atomi bomb survivors as alulated by the BEIR VII ommittee [7]. A linear, no-threshold fit and a linear-quadrati fit are plotted.

25 As shown in the atomi bomb survivor studies, inidene of aner varies greatly by organ, with the breast and thyroid showing the greatest exess relative risk (ERR) at a dose of 1 Sv. The overall ERR for solid aners is 0.63 per Sv: there was 63% greater inidene of solid aner in the exposed study group than the ontrol group [7]. After taking into aount many studies, the International Commission on Radiation Protetion (ICRP) developed a stohasti risk model that states that the absorption of 1 Sv of absorbed dose auses 5% inrease in aner risk [9, 20] Neutron Soures and Energy Nulear fission is the most ommon soure of neutron radiation. Fission ours when unstable nulei split into exited fragments, from whih neutrons evaporate as the fragments relax to lower energy states. Fission Fission ours when an atom aptures a neutron to beome an unstable isotope, and subsequently splits into two daughter atoms. These fragments are suffiiently exited to shed neutrons, beta partiles, and gamma rays, just as in spallation. Many neutrons are aptured either by other nulei, ontinuing the fission reation. In a reator, fission produts may be aptured by absorptive rods to prevent the reation from going out of ontrol. Fission reations produe a good deal of heat, requiring various dissipation proesses. Fast Neutrons Neutrons with energy greater than 1.0 MeV are haraterized as fast neutrons. Most neutrons produed in nulear proesses are fast neutrons. Fast neutrons travel with enough energy that the likelihood of them interating with other partiles is relatively low. Fast neutrons are not likely to be absorbed by a nuleus and ontinue a fission reation, for example.

26 In a fission reator, moderating materials are used to slow them and enable ontinuous reations. Water (in the form of H2O or D2O) and graphite are ommonly used moderators; the ollisions between the neutrons and the similarly-sized nulei are more likely to result in sattering than absorption. That is, the neutron is likely to remain free, but will be slowed by the ollision to thermal and epithermal energy levels, where their apture ross-setion is higher, allowing reations to ontinue. Epithermal Neutrons Many distintions are made in the range of energies between the fast and thermal neutrons. In general, neutrons with energy between 0.5 ev and 1.0 MeV an be haraterized as epithermal. The loser the energy of an inident neutron is to its surroundings, the greater its likelihood of being absorbed into a nuleus. Thermal Neutrons Thermal neutrons are in thermal equilibrium with their surroundings, having energies less than 0.5 ev. At a temperature of 290K, a ommon approximation of room temperature, the orresponding energy of a neutron is ev, so this is the value often used for alulations and simulations. Thermal neutrons have the greatest absorption ross setion. Fission reations require the presene of thermal neutrons, whih are absorbed by fissile nulei to start reations. Neutron Interation Cross Setion The probability of a neutron interating with an atom is dependent on the inident neutron s energy, and also on the atom s mirosopi neutron interation ross setion σ. This value, whih has units of either m 2 or barns (one barn equals m 2 ), represents the effetive ross setional area of an atom s nuleus, whih is where interation will our. σ = πr 2 (6) In this equation, r is the effetive radius of the atom s nuleus.

27 Depending on the material in question, a neutron may have variable likelihood to either satter or be absorbed by an atomi nuleus. With this in mind, atoms indeed have separate ross setions for sattering and absorption. The interation ross setion referred to here is the sum of the sattering ross setion σs and absorption ross setion σa. σ = σ s + σ a (7) On a marosopi level, the mirosopi ross setion σ is multiplied by the number density n of atoms in the sample in question, yielding a marosopi ross setion Σ: Σ = σn (8) In samples omposed of a mixture of elements, the interation ross setions ombine simply aording to the number density of onstituent atoms. Neutron Ativation Neutron ativation refers to the proess of exposing a material to a neutron beam, ausing the nulei to apture neutrons and beome unstable. This proess is highly useful for haraterizing the radiation emitted from a neutron soure and for identifying the elemental makeup of unknown samples. Neutron Soure Charaterization Neutron ativation an be used to measure the fluene rate of neutrons oming from a soure. A well-understood material an be exposed to a neutron soure for a period of time, ativating it. The ativity of the material an then be measured and related to the neutron fluene rate. The ativation of a sample is dependent on the fluene rate of the neutron soure, and also on the properties of the sample itself. A given nulide has a neutron apture ross setion, whih is a numerial representation of how likely it is to absorb a neutron. It is also important to take into aount the number of atoms present in a sample.

28 Of ourse a sample exposed to a neutron soure for a longer time will be more radioative, sine more interations are allowed to our. While exposed to a neutron soure, ativity in the target sample grows aording to a saturation fator S = 1 e λt (9) where λ is the deay onstant of the ativated target element, and ti is the time the sample is irradiated. One the sample is removed from the neutron soure, the ativated nulides will deay without new ones being made. The ativity of a sample annot be made while the sample is still being ativated, so this needs to be taken into aount. The deay ours exponentially, aording to a deay fator D = e λt d (10) In this equation, λ is again the deay onstant and td is the length of time the sample is allowed to deay. The Neutron Ativation Equation The equation relating eah of these fators is fairly straightforward: A = Nσφ(1 e λt i)(e λt d) (11) where A N σ φ λ ti td is the ativity of the sample is the number of atoms in the sample is the neutron apture ross setion is the fluene rate of neutrons is the deay onstant of the radioative isotope is the length of time of irradiation is the time of deay between ativation and ount

29 Neutron Ativation Analysis Neutron ativation analysis (NAA) takes advantage of the unique deay patterns of ertain ommon elements in order to identify them in samples of unknown material. While some elements are more detetable in trae amounts than others, up to 74 elements an be identified using NAA with a range of sensistivity between 1 and 1x10 7 piograms [10]. The nulei of different elements emit distint kinds of radiation as they transform bak to a stable form. For example, when naturally-ourring and stable gold-197 is exposed to a neutron soure, some atoms may absorb neutrons to beome gold-198. Gold-198 is unstable and most ommonly deays to merury-198 by emission of a beta partile of energy MeV and a gamma ray of energy MeV (Figure 5). Figure 5 - The deay sheme of gold-198 to stable merury-198 by beta emission. No other radioative element undergoes this same transition. Thus if radiations are olleted that indiate that these transitions took plae, it is likely that the original sample ontained gold-197, whih was ativated into gold-198 and subsequently deayed. By ativating an unknown sample and measuring its harateristi radiation (speifially the gamma), the elements in the sample an be determined, even when only trae amounts of the materials are present.

30 Computational Simulation of Radiation The use of radioative materials in any pratial appliation requires strit safety proedures and monitoring due to the potentially hazardous and damaging effets radiation an have on the environment. In plae of real-life experiments, omputer simulations an be used to generate useful information and results for even ompliated systems. Simulations have the further benefit of being highly flexible and ustomizable the geometry of an experimental setup an be adjusted easily, for example, or a hypothetial part an be modeled that would be impratial to raft for trial testing. The Monte Carlo Method In order to reate a notieable effet on the maro sale, the number of partile interations of a nulear proess must be very large. Furthermore, partile interations at the atomi level are entirely stohasti. The result is that partile transport an be modeled using a series of probabilities [26]. Monte Carlo N-Partile (MCNP) ode models the lifetime of numerous partiles in this way. A partile will be produed with harateristis (suh as type of partile, energy, and diretion) given by the input ode. After that, whenever there is a possibility for an event to our with that partile, the ode generates a random number and ompares it to known interation probabilities. These probabilities ome from the physial properties of the materials in question, suh as interation ross setions or deay onstants. The ode follows the initial partile and all subsequent interation produts aused by that partile until they are all absorbed as dose or exit the area of interest (Figure 6).

31 Figure 6 - A possible series of interations aused by an inident neutron and the order in whih they are traked [26]. By simulating a large number of partiles in this way, it is possible to obtain various types of usable information, inluding urrent aross a surfae, partile fluene rate aross a surfae, or energy deposition within a volume.

32 Methods 1. Monte Carlo (MCNP) Analysis i. Determination of Number of Histories The first geometry used was a series of five onentri one-entimeter-thik spherial tallies spaed six entimeters apart. The tallies were entered on a neutron point soure that produed neutrons of energy ev in random diretions. The world was filled with air with a omposition as defined by the National Institute of Standards and Tehnology: by weight, % arbon, % nitrogen, % oxygen, % argon (NIST). Figure 7 shows the spherial geometry. Figure 7 - The spherial geometry used in parts i. and ii. The thinner layers are the tally ells. The tallies used were F6 dose tallies, whih reord energy deposition averaged over the ell in units of mega-eletron volts per gram (MeV/g). The dose reorded by eah tally was not important at this stage, but F6 tallies would be used to full effet later, so it made sense to use them here. Also reorded is the tally error and a figure of merit, both of whih provide some insight into how statistially reliable the simulation is. The spherial setup would ensure that the maximum number of partile interations and the orresponding dose would be reorded at eah distane.

33 This spherial geometry was first used as a stage to determine the number of partiles to simulate in order to obtain reliable results. This means that the tallies pass the statistial heks built into MCNP, have a low tally error values, and whih have stable figures of merit as a simulation goes on. Seven simulations were run using the spherial geometry, with the number of partile histories simulated ranging from 10 to The results of these simulations an be found in Results setion 1.i. An example of the MCNP ode with a spherial geometry setup an be found in Appendix N. ii. Comparison of Geometries The experiments onduted at UMass Lowell would not have a spherial dosimeter setup; instead the dose readings would be taken by a dosimeter with a relatively small ross-setional area. In order to better understand the limitations of this setup, simulations with a linear setup of dose tallies were performed in MCNP for omparison with the spherial geometry used previously. In the linear geometry, five one-ubi-entimeter tally ells were plaed in a line and spaed six entimeters apart. A dis neutron soure of radius six entimeters was plaed six entimeters from the first tally ell and produed ev neutrons toward the tallies. As in the spherial geometry, the world was filled with air. The spherial geometry used in Part i., above, was used as the standard of omparison. Simulations with 10 7 partile histories were run and their results were ompared, with regard to absorbed dose and tally error. Figure 8 shows ross-setional views of both the spherial geometry and the linear geometry.

34 Figure 8 (A) The spherial geometry, with a entral point neutron soure. (B) The linear geometry, with dis neutron soure (line to the left of the squares). iii. Effet of Water Baking on Reorded Neutron and Gamma Dose from Thermal and Fast Neutrons MCNP was used to simulate thermal and fast neutrons and estimate dose for the thermal olumn and fast neutron faility at UMass Lowell. This would by simulated by setting up a simple neutron soure and plaing F6 dose tallies at set distanes from it. The tallies were set to reord dose from both neutrons and photons, whih would be seondary produts of partile interations. Cells ontaining water would be plaed just behind the tally ells, just as solid water bloks would be plaed behind dosimeters in the fast neutron dose experiment at the UMass reator. The thikness of these water ells ould be varied and the dose reorded by the tally ells ould be plotted and related to the thiknesses. The neutron soure was set at the origin of a 30 x 30 x 2200 m world filled with the same air used in previous simulations. The soure was a dis of radius six entimeters set to emit ev neutrons diretly down the x-axis, toward the tally ells. Based on the results of the previous simulations, it was deided that 10 7 partile histories would produe results with low error and stable figures of merit.

35 F6 neutron/gamma dose tallies were plaed at 200 entimeter intervals up to 2000 entimeters. The half-entimeter-square tally ells were baked by ells with two entimeter height and width ontaining pure light water. The thiknesses of the water ells were 0.1 m, 1 m, 4 m and 10 m for eah respetive simulation. Figure 9 shows a tally ell with a fourentimeter baking within the world ell. Figure 9 - A tally ell (green) with a 4m water baking (blue) within the retangular prism world ell (gray shade). Sine having multiple water ells in a row would ause umulative shielding effets, it was neessary to run many simulations: ten for eah water ell thikness, one at eah set distane from the neutron soure, with four different thiknesses, for a total of 40 runs. This was ahieved easily by utilizing the pstudy perl sript that an take one input file with multiple values given for a number of variables, separate it into the required number of standard input files, and then run them individually. The result of the F6 tallies were dose readings in units of MeV/g. A more ommon absorbed dose unit is the Gray, equivalent to 1 Joule per kilogram, or 6.241x10 9 MeV/g. The results from the tallies were onverted to units of Gray for analysis. The same simulation was run with the dis neutron soure produing fast neutrons of energy 1 MeV.

36 The results of these simulations an be found in Results setion 1-iii. The full pstudy input files an be found in Appendix N. 2. Fast Neutron Dose Upon arrival at the UMass Lowell reator faility, the Landauer Luxel+ dosimeters (Figure 10) were sorted aording to how they would be used. The dosimeter labeled Test was left in an offie outside the reator area to at as a ontrol sample and provide a baseline for bakground radiation. Figure 10 - A Luxel+ dosimeter, with pen for sale. The remaining dosimeters were labeled 1 through 10 and were exposed to radiation from the reator ore in pairs as shown in Table 4. The ninth and tenth dosimeters were left in the ontrol room in the reator area for the duration of the experiments and ould have been used as bakups if any stage of the experiment was disrupted.

37 Table 4 Reator power and dosimeter exposure times for the fast neutron irradiation experiment. Dosimeter Number Reator Power Output (watts) Duration of Exposure (minutes) 1, , , , Figure 11 shows two external views of the bunker wherein the dosimeters were exposed to the reator radiation. Figure 11 - (Left) An overhead view of the neutron failities. The thermal olumn and fast neutron bunker are both visible. (Right) Another external view of the fast neutron bunker. High-density onrete and other insulating materials are used to ontain the radiation released within. A ollimation tube, visible in Figure 12, onneted this room to the reator. For eah stage of the experiment, the dosimeter pair being used, baked by solid water, would be positioned diretly in front of the tube. The dosimeters being used were held in a plasti bag that was affixed to the solid water. While the UMLRR fast neutron irradiation faility has a uniform fluene rate distribution within 10% of the average [24], the dosimeters would get the best results if they were plaed in the enter of the partile beam, where fluene is most uniform. The bags ontaining the dosimeters were positioned onsistently, using marks on the solid water bloks, to ensure that the dosimeters were in the same loation relative to the tube for eah stage of the experiment.

38 Figure 12 - A piture of the setup inside the bunker. The movable rak system (holding the solid water, right of piture) ould be moved from outside the bunker to the front of the ollimation tube (left of piture). The tube has a 30m diameter In order to position the dosimeters in front of this tube without entering the bunker while it was being irradiated, a movable rak system was used (Figure 13). The 18 entimeters of solid water were set up on a platform whih ould be moved by a pulley system between the proper position in front of the ollimation tube and a loation outside the bunker where dosimeters ould be affixed to the setup safely. This also made it possible to exhange the dosimeters while keeping the reator powered up, sine raising and lowering the power for eah pair of dosimeters would have made the experiment take muh longer. A Geiger-Müller ounter and a neutrondeteting variant were set up in this area to give indiations of the radiation levels there. Figure 13 - The solid water positioned in front of the ollimation tube. The bag ontaining the dosimeters is held to the solid water bloks by tape.