Keywords

3.1 Beam Loss in Particle Accelerators

Ideally, a particle accelerator works with 100% efficiency, transporting all charged particles from a source to a target while giving them the required total energy. In reality, collisions between beam particles and matter are mostly unwanted, but unavoidable. With increasing energy and intensity of the particle beams delivered, the effects of particle losses (“beam loss”) are becoming more pronounced. As an example, the CERN Proton Synchrotron (PS) accelerated in the early 1960s proton beams with an intensity of 1012 particles per pulse. In the years around 2000, the beam loss at the extraction septum towards the Super Proton Synchrotron (SPS) exceeded this value by a factor of 2–3 [1].

Sources of unwanted beam-matter interactions, leading as a result to deviations from the ideal orbit, are [9]:

  • Imperfect beam line vacuum making beam particles collide with rest gas molecules. This interaction deviates the beam particle from its longitudinal course and may break-up the rest gas particle.

  • Electromagnetic fields induced by high-intensity beams increase the transverse momentum of beam particles, and lead to their eventual loss by collision with the beam tube.

  • Beam steering elements, such as kickers and septa. Some particles inevitably impinge on the blade, a thin foil in a septum defining the electric field which brings the particles on a new course. Another loss mechanism is misfiring of a kicker’s electromagnetic circuits, the “kicked” particles will go astray and collide with the beam line at positions which can be predicted from beam dynamics.

  • Close encounters of colliding beams in the centre of a physics detector, inducing transversal momentum, leading to small deviations from the “perfect” particle orbit. In circular colliders, these minute deviations may lead to the loss of the particles many turns after the interaction.

In Sect. 2.6, controlled interactions with the beam were introduced:

  • Collimators, skimming particles with transversal excursions at predefined locations.

  • Beam-dumps, absorbing the particle beam after the target or after it has become too diluted to be useful for further exploitation in a collider.

  • Finally, fixed targets in physics detectors for extracted beams are obviously sources of beam-matter interactions.

The first two beam loss mechanisms are distributed over the whole length of the accelerator while all others are localised.

A rare hazard is the direct exposure of personnel to the particle beam. In modern accelerator facilities the risk of this happening is reduced to near zero by accelerator safety systems and access control systems, described in Sect. 5.3. These systems make it near-impossible for a person to enter the accelerator area when particle beams are produced.

3.2 Beam-Matter Interaction

Particles lost from the beam collide with material from the accelerator. Given sufficient energy, they traverse the beam line vacuum tube, the magnets and other equipment. Their path ends in the shielding walls or in the ground or air surrounding the accelerator. This section describes the passage of particles through matter. The multiple interactions between beam and matter generate the radiation fields at an accelerator during operation. They are also responsible for activation of the material. Here, a qualitative overview of this complex subject is given, without going in the physical details and their mathematical description. For more detailed descriptions, one can refer to [3, 6, 7].

3.2.1 Electrons and Positrons

The following interaction mechanisms contribute to the energy loss of electrons and positrons in matter, and to the emission of ionising radiation.

  • Ionisation. An inelastic collision of the incoming electron with a bound electron mobilises the latter as a so-called δ-electron. The δ-electrons can in turn ionise further atoms on their path.

  • Bremsstrahlung. The electron emits X-ray photons during deceleration or deviation.

  • Pair production (PP). A photon with an energy E > 2 m0c2 = 1.022 MeV can convert to an electron-positron pair in the electric field of an atom.

  • Positrons at rest annihilate with an electron into two or three photons.

Low-energy electrons interact predominantly by ionisation. At the critical energy Ec, the probability for ionization and Bremsstrahlung are equal, at higher energy, Bremsstrahlung prevails. The value of critical energy is approximatelyEc(MeV) = \( \frac{800}{Z+1.2} \), where Z is the atomic number of the target material [10].Radiation length Xo is the depth in matter in which Bremsstrahlung has reduced the energy of a high-energy electron (E >> Ec) to 1/e of the original value. The energy dependence of the different interaction mechanisms is illustrated for lead in Fig. 3.1.

Fig. 3.1
figure 1

Fractional energy loss per radiation length in lead for each of the interaction mechanisms as a function of electron or positron energy. The critical energy Ec is situated where the curves for Ionization and for Bremsstrahlung cross. (From [8])

Full size image

In thick absorbers, electrons or photons with an energy well above the critical energy Ec initiate an electromagnetic (EM) shower or cascade. In an EM cascade, the interaction mechanisms of compton scattering, pair production and bremsstrahlung alternate and lead to several generations of electrons, electron – positron pairs and high-energy photons. The kinetic energy of the particles decreases in each generation. The cascade stops propagating once the average energy of the electrons has fallen below the critical energy Ec. From this moment on, the probability that the electrons ionise atoms is higher than that of Bremsstrahlung emission. The cascade loses one of its driving particles because the electrons now dissipate their energy preferably by ionisation and come quickly to rest. A schematic illustration of an EM cascade is given in Fig. 3.2.

Fig. 3.2
figure 2

Schematic development of an electromagnetic cascade from a high-energy electron (E> > Ec). Electron tracks are blue, positrons red and photons green. Interactions: B Bremsstrahlung emission, P pair production, C Compton scattering. (Image by the author, after [10])

Full size image

3.2.2 Protons and Charged Heavy Particles

Protons and charged ions traversing matter are subject to the following interaction mechanisms:

  • Electromagnetic interaction with target electrons. Charged particles interact with the atomic electrons and nuclei of the target material. Electronic energy loss is the most probable mechanism, because of the mass difference between the projectiles only small amounts of energy in the range of eV can be transferred. This energy leads to excitation of the target atoms and electrons in rotational and vibrational states and finally to heating of the material. Target atoms may also be ionised by the ejection of a δ-electron.

  • Electromagnetic interaction with target nuclei. These interactions are less frequent than with electrons, but more energy is transmitted. The results of these collisions are a change of direction of the projectile and possibly a displacement of the target atom.

  • Nuclear interactions with target nuclei. Once a positively charged projectile overcomes the Coulomb-barrier of the nucleus, (typically with a kinetic energy of a few MeV) it can penetrate the nucleus and trigger nuclear interactions leading to a rearrangement of its constituents. Excess energy is carried away by protons, neutrons, or small nuclear fragments (nuclei of 2H, 3H, 3He, 4He, …) leaving the nucleus. This is called the spallation process . The kinetic energy of the spallation nuclei and fragments depends on the kinetic energy of the projectile and the released binding energy of the nucleus. Energetic spallation nuclides trigger further electromagnetic and nuclear reactions and a hadronic cascade may be the result.

  • Hadronic interactions in target nuclei. When the energy of the projectile is high enough, new particles can be generated in nuclei. Pions are produced at energies from 290 MeV on, other, heavier hadrons require correspondingly more energy. The generated particles are unstable, and they decay into stable particles. Charged pions decay to muons and neutrinos and neutral pions to photons. Together with the spallation nuclei and fragments, all these particles form part of the hadronic cascade.

3.2.3 Neutrons

Neutrons are produced at accelerators in two processes:

  • Nuclear and hadronic interactions of charged particles. As seen above, neutrons are emitted in the spallation process and from hadronic interactions in target nuclei.

  • Photonuclear reactions . As the name suggests, photons can trigger nuclear reactions, which often result in the emission of a photoneutron . The effect is enhanced for heavy target materials with high atomic number Z. In hadronic cascades, and thus in proton or ion accelerators, their rate is negligible against neutron production in the hadronic cascade. In electron accelerators with primary energy above 15 MeV, the production of photoneutrons cannot be neglected. This includes electron linacs for radiotherapy, where heavy metal beam collimators are a source of photoneutrons. They must be considered in the shielding design of such accelerators and they contribute to the unwanted dose in the patient, which is not directed to the tumour.

Free neutrons have a lifetime of about 15 min, they are neutral and can traverse thick shielding walls, and they have a spectrum of effects which spans many orders of magnitude in neutron energy. While charged particles and photons are efficiently attenuated by heavy materials, neutrons are dominating radiation dose and effects outside of particle accelerator shielding (Fig. 3.3).

Fig. 3.3
figure 3

Components of a secondary particle cascade developing in concrete. The primary particles are electrons of 1 GeV, triggering an EM cascade (left column) or protons of 1 GeV, triggering a hadronic cascade (right column). The plot shows the fluence of secondary particles released by 109 primary particles hitting the wall from the left. Note that the fluence of secondary particles in a hadronic cascade is several orders of magnitude higher than in an EM cascade. (Image by the author)

Full size image

3.2.4 Radiation Damage

Radiation damage summarises several effects which are changing the properties of material struck and traversed by radiation or particle beams [2, 5]:

  • Electromagnetic interactions of projectiles with the target nucleus can lead to atomic displacement. In solid materials atomic displacements lead to crystal defects. A high density of displacements may lead to embrittlement, altering the mechanical resistance of the material. This is of primary concern in high-intensity production targets (Sect. 2.6.2).

  • Nuclear interactions with target nuclei lead to its nuclear transmutation. The electronic structure of the produced atom does no longer match the crystal lattice and represents another type of lattice defect.

  • Heating occurs in target materials struck by a particle beam, because the kinetic energy deposited by interactions of the projectile with the target electrons cannot be removed immediately. In extreme cases, local melting may occur, leading eventually to the destruction of the equipment struck by the beam.

These effects are noticeable at places in the particle accelerator where beam loss is prominent: in beam dumps, targets and collimators (Sect. 2.6). Other equipment are the electromagnetic septa used to change the direction of particle beams. These components must be built robust enough to withstand the effects of particle impact. Sometimes, they are built to survive a certain amount of beam before they are preventively exchanged.

Electronic components suffer from three effects mediated by particle beams and hadronic or EM cascades [11]:

  • Single event effects have their origin in the deposition of a minute amount of charge in an integrated circuit (IC). The charge may flip a logical gate in the IC and provoke an unexpected behaviour. This is a stochastic effect and its probability is proportional to the fluence of secondary particles. Single-event effects are prominent in highly integrated microelectronics, where very small amounts of charge are sufficient to provoke a change.

  • Total Ionisation dose is the continuous accumulation of absorbed dose in semiconductor materials. Ionisation leads to the creation of electron-hole pairs. In conducting materials, they rapidly recombine, but in the insulating oxides of a semiconductor, holes may persist and, at higher density, permanently change the electronic characteristics of the material.

  • Displacement damage in silicon, mainly provoked by neutrons. Displaced silicon atoms leave gaps in the crystal lattice, changing its electronic properties.

3.2.5 Activation of Matter

Energetic protons cause spallation reactions, having as products nucleons, light nuclear fragments, and a remaining rest nucleus, unstable most of the time. Electron beams can activate matter by photonuclear reactions. The previously non-radioactive material has become radioactive by the exposure to the particle beam.

Estimates for activation are difficult to make, they depend on the energy of the projectile, and the composition of the target material. Rules-of- thumb and approximate formulas are given for electron accelerators by [10] and for proton- accelerators by [34] and [33]. Activity estimates with higher precision are possible with Monte-Carlo radiation transport programs coupled to a nuclear model. Their activation estimates compare with experiment within a factor of two for “good cases” [4] and may be off by an order of magnitude for other nuclei. The prediction of dose rates from activated material by this method is often acceptable because of cancellation effects between nuclei with over- and under-estimated activity.

A frequently repeated rule-of-thumb states that “hands-on” maintenance at a particle accelerator is possible, if the power of beam loss (the product of particle energy and current of lost particles) does not exceed the value of 1 W per metre. This “rule” for which the original source could not be identified, must be applied with care. It originates in the 1960s or 1970s and experience shows that the resulting activation and dose equivalent rate of the material is frequently too high according to modern standards.

3.3 Ionising Radiation

The term “ionising radiation” assembles several physical phenomena that have in common that they can ionise matter. The warning signs against ionising radiation contain a stylised image of a radioactive source (Fig. 3.4).

Fig. 3.4
figure 4

Warning sign against ionizing radiation and marking of radioactive material transport, after [13], [29]. (Image source: https://publicdomainvectors.org)

Full size image

3.3.1 Types of Ionising Radiation

Two main families of ionising radiation can be distinguished: directly ionising radiation, mediated by charged particles, and indirectly ionising radiation, by neutral particles and photons.

3.3.1.1 Directly Ionising Radiation

Directly ionising radiation consists of charged particles. They collide with atoms and ionise them by the electromagnetic interaction. One characteristic of directly ionising radiation is that they have a definitive range in matter. Directly ionising radiation may consist of

  • alpha (α) radiation, helium nuclei emitted by heavy radioactive nuclei,

  • beta (β) radiation is a manifestation of the weak interaction in unstable nuclei, which de-excite under emission of an energetic electron or positron.

  • In particle accelerators, other directly ionising particles are produced in collisions between particles and matter, for example muons, pions and protons.

3.3.1.2 Indirectly Ionising Radiation

Indirectly ionising radiation does not interact directly with the atomic electrons. In an intermediate step a charged particle is generated (often a secondary electron) which in turn ionises matter. Due to this interaction mechanism, indirectly ionising radiation is exponentially attenuated in matter and has no finite range. One distinguishes between two fundamental families:

  • X- and gamma (γ) rays are energetic photons, quanta of the electromagnetic radiation with energy E > 1 keV. Photons interact with atoms by the photoelectric effect and the Compton effect, releasing a secondary electron. This secondary particle imparts its energy by electron-atom collisions to matter. Photons with very high energy (E > 6 MeV) can interact with the atomic nucleus (photonuclear effect). The excited nucleus will de-excite by emitting either other photons or particles or, for heavy elements, undergo a fission reaction (photofission).

  • Neutrons (n) are neutral, nuclear particles emitted by some heavy nuclei, or released in particle – matter interactions. Neutrons interact with atomic nuclei, either scattering elastically (preservation of total kinetic energy, i.e. the internal state of the target nucleus remains unaffected) or inelastically (nuclear reactions take place in the target nucleus). Charged secondary particles are emitted by the excited target nucleus, and a recoil nucleus remains, moving with an energy determined by the reaction kinetics. Recoil protons can take a large fraction of the neutron’s energy, because they have nearly the same mass. Certain nuclides absorb thermal neutrons having energies below 0.025 eV. They become activated, with half-lives ranging between fractions of a second and many decades of years.

3.3.2 Sources of Ionising Radiation at Accelerators

Due to the high concentration of energy available in accelerated particles, a large variety of ionising radiation types with wide energy distributions can occur at accelerators. This distinguishes accelerators from facilities in the nuclear power generation cycle, from non-destructive testing and from (most) medical applications.

Two phases can be distinguished with respect to ionising radiation in a particle accelerator:

  • During accelerator operation, beam loss, beam-target and beam-beam collisions lead to radiation spectra with an exceptionally wide range of particles and energies, elsewhere encountered only in cosmic radiation. Ionising radiation produced in particle-matter interactions during accelerator operation is also named prompt radiation.

  • During maintenance phases, personnel may be exposed to activated material and, exceptionally, to radioactive contamination.

3.3.2.1 Prompt Radiation

At electron accelerators, prompt radiation originates from the electromagnetic cascade, consisting of electrons, positrons, and photons. If the energy of the primary electron is below E < 10 MeV, then only these particles contribute to the radiation field. This is the case in medical linacs in diagnostics and therapy in a low-energy setting. Once the threshold of approximately 10 MeV is passed, Bremsstrahlung photons can trigger reactions in nuclei in the so-called photonuclear effect, with production of neutrons. At higher energies, hadrons accompany the EM cascade.

In proton accelerators, the prompt radiation field is described by hadronic cascade, with protons, neutrons, photons, electrons, and positrons as the prevalent components of the radiation field.

An important characteristic of prompt radiation is, that it ceases with the stop of the accelerator. In a properly built accelerator facility, the prompt radiation is correctly shielded by constructing the accelerator underground, or by protecting it with shielding walls made from iron and concrete. It represents in general no risk for workers or the population.

3.3.2.2 Radiation from Activated Material

Ionising radiation emitted from activated material is characteristic for the radionuclides generated, it consists mainly of photons and electrons, with kinetic energies below a few MeV. Different from prompt radiation at accelerators, it does not cease with the stop of the accelerator. Since prompt radiation from an operating particle accelerator is well shielded, exposure to ionising radiation emitted by activated material constitutes the largest contribution to personal dose at an accelerator.

Radiation dosimetry and radiation protection against the exposition to activated material use the same detectors, dosimeters, and procedures as the well-developed field of radiation protection in nuclear or medical facilities.

A particular type of activated material are radiopharmaceuticals, which are produced in small accelerators from stable isotopes by proton bombardment. The produced radionuclides are collected within targets or on foils and they often represent a radioactive contamination hazard. This describes the danger of transferring loose radioactive materials from surfaces to the body and of eventually ingesting or inhaling them. This hazard can be prevented by application of industrial hygiene measures, as they are also employed in workplaces with chemical hazards: protective clothing, Respiratory protection with filters or active air supply. The relative advantage of radioactive contamination over chemical hazards is, that ionising radiation can be detected more easy than chemical contamination.

3.4 Radiation Dosimetry at Accelerators

3.4.1 Dose and Dose Equivalent

The measurement of the detrimental effect of ionising radiation is based on a physical quantity, absorbed dose D . It is defined as the ionising energy imparted on a small target of mass δm:

$$ D=\frac{\delta \epsilon}{\delta m} $$

In the SI system, the quantity absorbed dose D is measured in the unit J kg−1, which receives the special name Gray (Gy).

One can distinguish two effects of ionising radiation on humans:

  • Tissue reactions were previously called deterministic effects [15] because they appear once a subject has been exposed to an absorbed dose above a threshold. The mildest form is reddening of the skin (erythema), from approximately 0.5 Gy applied to a limited portion of the skin. The thresholds for tissue effects are rather high and they play a secondary role in occupational radiation protection.

  • Stochastic effects of ionising radiation: these are caused by modifications of the genetic material in cells, either by direct interaction with ionising radiation, or by changes in the cell medium impacting. The probability of developing cancer increases with the radiation dose received, without lower threshold for the effect. The effect is cumulative for chronic exposure.

After World War 2, systematic health studies of survivors of the nuclear bomb explosions over Hiroshima and Nagasaki started, complemented by the follow-up of patients having received radiation treatment. The epidemiological observations are combined with continuously refined biophysical models of exposure to ionising radiation. This work culminated in the recommendations of the International Commission on Radiological Protection (ICRP), with ICRP Publication 103 [14] the most recent in the series.

In the present discussion, tissue effects are neglected, their occurrence at accelerators is extremely rare. In its recommendations, ICRP defines a protection quantity for stochastic effects of ionising radiation. It is based on absorbed dose D, but it takes into account the difference in radiation sensitivity of tissues and organs in the body, and the effectiveness of different radiation types to cause cancer. This leads to the quantity effective dose E. Effective dose E is not a purely physical quantity, but it includes a measure of the probability to develop cancer after being exposed to ionizing radiation:

$$ E=\sum \limits_T{w}_T\sum \limits_R{w}_R{D}_{T,R} $$

In this formula, DT,R is the absorbed dose in averaged over a tissue (organ) T by radiation type R. wR and wT are the radiation- and tissue weighting factors, respectively. Effective dose, by definition, is an average of dose over the whole body and cannot be measured directly.

The sum \( \sum \limits_R{w}_R{D}_{T,R} \) is called equivalent dose, it accounts for the radiation effect on a single tissue (organ) T. The physical quantity to express equivalent dose and effective dose is J kg−1. To distinguish it from absorbed dose, it receives the special name Sievert (Sv).

The paradigm of radiation protection is, that the probability of radiation detriment to a subject is proportional to the amount of effective dose E received, without a lower threshold. Radiation detriment includes the probability of developing cancer and the ensuing loss of years of life and quality of life. Effective dose limits, limiting the probability of the occurrence of stochastic effects, are determined by comparison with the risk of incapacitating or fatal accidents in other industries. An acceptable level of risk is determined and expressed in effective dose E, which is periodically reviewed by the ICRP. Limits are also set to prevent tissue reactions, with a large safety margin below the threshold dose for reactions (Table 3.1).

Table 3.1 Dose limits for ionizing radiation [14]
Full size table

To provide measurable quantities, the International Commission on Radiological Units and Measurements (ICRU) has defined so-called operational quantities, approximating effective dose. They receive the generic name dose equivalent with symbol H and, like effective dose, are quantified in the SI system with the unit J kg−1 with the special name Sievert (Sv). The ICRU introduced several operational quantities with specific definitions to cover different exposure situations, for example of the eye lens or the skin. The most commonly used operational quantities are those for the exposure of the full body with Hp(10) (spelled “H-p-10”) for the calibration of personal dosimeters and H*(10) (“H-star-10”) for survey instruments.

In contrast to effective dose E, dose equivalent quantities are defined in such a way that they can be realized in standard laboratories and measured with calibrated instruments, within uncertainties.

3.4.2 Practical Radiation Dosimetry at Accelerators

The radiation fields around a working particle accelerator are consisting of different types of ionising radiation, extending over broad ranges of energies. This makes dose equivalent measurements at accelerators challenging, and special instruments have been developed over the years to cope with the characteristics of these fields. In shutdown periods, when the accelerator is stopped, the origin of radiation is activated material, containing relatively long-lived radioactive nuclei (with half-lives of days to years), created in beam-matter interactions during the preceding operation periods of the accelerator. These nuclei emit gamma (photon) and beta (electron/positron) radiation. Dose equivalent of these radiations can be measured with standard radiation protection dosimeters and survey instruments, as developed for dosimetry in nuclear and medical radiation facilities.

In a radiation protection program at an accelerator facility, radiation measurement is essential for the following tasks

  • measuring dose equivalent at workplaces to make a prospective assessment of working conditions, and to demonstrate the adequacy of the radiation shielding put in place;

  • monitoring the dose equivalent to personnel, to demonstrate compliance with legal dose limits;

  • monitoring radiation emitted to the environment, to demonstrate compliance with emission and immission limits (Sect. 4.6.1).

Dosimeters are used to measure dose or dose equivalent. In this section, the term dosimeter is used likewise for passive and active devices whose purpose is the measurement of dose equivalent. Based on their area of usage, they are may be personal dosimeters, radiation survey instrument or radiation monitors.

One distinguishes between passive and active dosimeters.

  • Passive dosimeters are based on a physical effect which is (roughly) proportional to dose equivalent. The magnitude of the physical effect creates a proportional signal once they are read-out after a defined period of exposure. They are used when the immediate display of a result is not essential, for example for personal dosimetry at workplaces with a moderate dose equivalent rate, or for long-term monitoring of the environment.

  • Active dosimeters have a detection mechanism which can be directly converted to an electrical signal. Active dosimeters have the advantage to give an immediate display of the dose equivalent, but they are usually larger and more expensive than passive dosimeters. They are used in the form of electronic personal dosimeters to monitor workers at workplaces where the dose equivalent rate is so high that they could accumulate a significant fraction of the dose limit in a short time. Active dosimeters are also used for radiation surveys, and as alarm monitors where sudden increases of the radiation intensity are possible and Survey instruments and active radiation monitors usually indicate the instantaneous dose equivalent rate \( \dot{H}=\raisebox{1ex}{$ dH$}\!\left/ \!\raisebox{-1ex}{$ dt$}\right. \).

Many detectors based on different physical principles exist for the quantification of ionising radiation. Only few of them can be employed as a radiation dosimeter. To make a good dosimeter, the energy-dependent response of the detector to a physical quantity describing the radiation field (for example, particle fluence) must be approximately proportional to the energy dependence of the operational quantity to be measured, for example H*(10). Hardly any physical radiation detector shows spontaneously the required proportionality of the physical effect to dose equivalent to be used as a radiation dosimeter. It is necessary to modify the energy response of the physical detector to mimic the energy dependence of the dose equivalent quantity to be measured.

This section can only give a very short overview about radiation detection and dosimetry. [20] or [19] give a more detailed treatment of radiation detection, whereas [22] is still a standard reference for radiation dosimetry. [12] gives a modern account of this subject.

3.4.2.1 Photon (Gamma) Dosimeters

The largest fraction of dose equivalent to which workers are exposed in accelerator facilities comes from photons and is accumulated during shutdown periods when the accelerators are maintained. The source of the photons is material activated by interactions with particles during the preceding operational periods. The isotope content of activation products in accelerator facilities is different from those in the nuclear industry by type and concentration, but the same methods of detection and measurement can be employed. Long-lived activation products emit gamma (photon) and beta (electron or positron) radiation.

Photons are indirectly ionising particles, as a first step to detection a secondary electron must be created by the photoelectric effect or by Compton scattering. Most active photon dosimeters rely on one of two detection mechanisms: gas ionisation and creation of free charges in semiconductors. Passive photon dosimeters use thermoluminescence (TL) and optically stimulated luminescence (OSL) as detection principles. Photographic film has been replaced by these technologies in most applications.

3.4.2.1.1 Gas-Filled Photon Detectors

Three different types of gas-filled detectors for photons are used in radiation protection: ionisation chambers, proportional counters, and Geiger-Müller counters. The commercially available types have in common a cylindrical geometry with a collection anode in the centre (in proportional counters and GM counters, an anode wire) (Fig. 3.5). Secondary electrons are created in the detector wall and registered in the filling gas by ionisation. The average energy to create an electron-ion pair in gases, the ionisation energy, lies between 20 eV and 40 eV, for air it is approximately 32 eV. The charges are separated by an electrical field between the anode and the chamber wall, and electrons drift to the central anode. The chamber constitutes a cylindrical capacitor for which the radial component peaks at the anode wire.

Fig. 3.5
figure 5

Schematic construction of a gas-filled detector for ionising radiation. (Adapted from [20]. Here, the signal is coupled out with a high-pass filter, in DC ionisation chambers, a low-pass filter would be employed, by changing the position of capacity and resistor to ground)

Full size image

The three types of gaseous detectors can be distinguished by applied voltage V0 between chamber wall (cathode) and central anode:

  • Ionisation chambers for radiation protection operate usually in DC mode, with a low-pass filter connected to the anode. The applied voltage V0 is in the leftmost region in Fig. 3.6, high enough to separate the created electron-ion pairs before they can recombine. The ionisation current comes from the electron-ion pairs created by the secondary electron along its path and is proportional to the energy deposited in the gas by the incident radiation.

  • In proportional counters, the field strength E(r) close to the wire is high enough to ionise additional atoms by the secondary electrons. A charge avalanche with an amplification factor A ≈ 102⋯106 will result. In a certain voltage range, depending on the gas type and pressure and chamber geometry, the amplification factor A is independent of the energy of the incoming particle, and the thus signal amplitude proportional to particle energy. This is the proportional regime.

  • In a Geiger-Müller counter, the electrical field close to the anode wire is so high that the charge avalanche triggers further avalanches all along the wire and depletes the counting gas from neutral atoms. In this regime, the amplification factor is A ≈ 1010 and independent from the energy of the incoming particle.

Fig. 3.6
figure 6

Signal height collected versus applied voltage in gas-filled radiation detectors with central wire. (After [20])

Full size image

The energy dependence of the three gaseous detector types is mainly determined by the probability that a photon interacts in the wall and emits secondary electrons into the counting gas. Ionisation chambers with walls made from tissue equivalent (TE) plastic and filled with TE gas give a signal proportional to absorbed dose in tissue, which is a good approximation of dose equivalent for photons and electrons. For other particles, a so-called quality factor must be applied. In instruments based on Geiger-Müller and proportional counters, metallic filters are placed around the counter chamber to influence the energy dependence of their response to photons. This strategy is successful in a limited energy range, usually between 50 keV and 1.3 MeV or 3 MeV. Below this interval, the dosimeters are insensitive (low energy photons are absorbed ion the detector wall), above they have a strong overresponse.

Proportional counters and Geiger-Müller counters are light and small, their sensitivity is ideally suited to assess photon radiation from activated accelerator components. Ionisation chambers can be used to monitor the radiation emitted by an accelerator in operation. For this, their factory calibration to reference photon sources, must be adapted to take account of the mixed radiation field.

3.4.2.1.2 Thermoluminescence and Optically Stimulated Luminescence

Thermoluminescence (TL) and optically stimulated luminescence (OSL) are based on a similar physical effect. Detectors based on this effect are at the basis of most modern passive photon and electron dosimeters. TL and OSL are made from crystalline materials with defined energy bands for electrons, for practical applications they are ground to powder and pressed in tablets or rods (Fig. 3.7). Ionising radiation creates free charges in the crystal’s conduction band. Instead of immediately returning to the valence band under emission of light, some electrons are “captured” in so-called trap levels, situated energetically below the conduction band. The trap-levels are metastable states with a long lifetime, and thus capable to store information about the radiation: the number of occupied trap levels is proportional to the dose absorbed in the TL or OSL crystal.

Fig. 3.7
figure 7

Thermoluminescence material crystal powder, disks and rods of different size, with permission from [21]

Full size image

To read out the stored information, one stimulates the electrons in the trap levels either thermally by the application of heat (TL), or optically with a laser of a defined wavelength (OSL). The excitation energy moves electrons from the trap levels to the conduction band, from where they return rapidly to the valence band under emission of light. The photons emitted by the TL or OSL material are amplified by a photomultiplier, the integrated light intensity is proportional to the dose absorbed in the material.

A commonly employed TL material is LiF. Its average atomic number is tissue equivalent. TL detectors from LiF in which the isotope 6Li is enhanced are used as thermal neutron detectors in passive neutron dosimeters (Fig. 3.8).

Fig. 3.8
figure 8

Different TLD holders to be inserted in a filter cassette, providing the correct energy response, with permission from [21]

Full size image
3.4.2.1.3 Semiconductor Detectors

A semiconductor is a crystalline solid from Silicon (Si) or Germanium (Ge), in which the atoms are arranged in a regular, spatially repetitive pattern. The base materials are usually doped with elements in the neighbouring columns of the periodic table of elements. In semiconductor crystals, the highest energy level of the valence band is separated from the conduction band by a gap with a width of a few eV. At room temperature, a few electrons occupy the conduction band and the semiconductor exhibits a small dark current. This can be eliminated by cooling the detector with liquid nitrogen to T = 77 K.

The working principle of a semiconductor detector is to release an electron from the valence band and transfer it to the conduction band. The energy required for this process, the ionisation energy is composed of the energy to create an electron-hole pair and the energy required to traverse the band gap. It is of the order of a few eV. The hole signifies an empty place in the valence band. The ionisation energy of a semiconductor diode is much smaller than the ionisation energy in a gas-filled detector and for an identical photon energy, more charges are produced in a semiconductor detector than in a gaseous detector. This results in a better energy resolution. The detector also has a higher atomic number (Si: Z = 14; Ge: Z = 32) and a higher density than gases, these two factors increase the cross section for the photo- and Compton effects and thus the chance to absorb a considerable part of the gamma photon’s energy. Such detectors from pure Germanium or, more rarely, pure Silicon are used in radiation protection as spectrometers for the identification of radionuclides by their characteristic emissions.

For the application as a dosimeter, positively and negatively doped semiconductor layers enclose a layer of intrinsic (undoped) material to form a PIN diode. A voltage is applied across the PIN junction to empty the intrinsic layer from all charge carriers (reverse bias). When ionising radiation traverses the intrinsic layer, electron-hole pairs are created and separated by the bias voltage, generating a small charge pulse which is amplified for detection.

The small size of PIN diodes and the possibility to power the bias voltage and the detection circuit with small batteries have made PIN diodes the detectors of choice for personal electronic dosimeters, and many variants of such devices are on the market.

3.4.2.2 Neutron Dosimeters

Like photons, neutrons are indirectly ionising particles. Two physical processes are employed in neutron dosimeters for the detection of thermal neutrons and energetic neutrons.

3.4.2.2.1 Thermal Neutron Detectors and Rem-Counters

Thermal neutrons (by convention, neutrons with a kinetic energy of less than 0.025 eV) can be captured by specific nuclei, such as 3He, 6Li, 10B, and 113Cd. After absorbing the thermal neutron, these nuclei become unstable and disintegrate into several easily detectable charged particles. Passive thermal neutron detectors are LiF TL detectors enriched in the isotope 6Li. A second detector of the same size, enriched in 7Li, can be used to estimate the photon component in the radiation field for background subtraction. Active detectors are proportional gas counters filled with either 3He or 10BF3 (a molecule containing boron, and gaseous under normal conditions). Neutrons with higher than thermal energies t must be slowed down in a moderator to be captured with good efficiency in a thermal neutron detector.

Neutron moderators are made from materials which are rich in hydrogen, mostly polyethylene (PE, CH2). In this material, fast neutrons are slowed down in multiple scattering events with protons. In a rem-counter, (from Rem–Roentgen equivalent man, an obsolete unit for an equally obsolete, pre-1990 dose equivalent quantity), the 3He or 10BF3 counter is surrounded by a composite moderator, consisting of PE to moderate and borated plastic to absorb surplus thermal neutrons. (Fig. 3.9). The moderator layers are arranged in a way to model the energy dependent response of the instrument as closely as possible to the fluence-to-dose rate conversion coefficient for neutrons. A rem-counter loses sensitivity for neutrons with energies of more than a few MeV. They are insufficiently moderated by the PE layer and have a small detection cross section in the thermal neutron detector. These rem-counters can nevertheless be used in locations where one has previously performed an in-situ calibration: one determines dose equivalent from high-energy neutrons with an independent method and modifies the calibration coefficient of the standard rem-counter so that is shows the same result. This method requires that the ratio of low- to high-energy neutrons in the radiation field remains approximately independent of the details of accelerator operation, which must be verified independently. As a rule of thumb, a standard rem-counter shows approximately half of the correct dose equivalent when used to verify the shielding of high-energy accelerators.

Fig. 3.9
figure 9

Different commercially available rem-counters. (From [16])

Full size image

An independent measurement of the full neutron dose equivalent can be made by a so-called extended range rem-counter. In this instrument, heavy metal layers complement PE and borated plastic. They increase high-energy sensitivity by neutron spallation. The first documented exemplar of an extended range rem-counter can be found in [17], today a few companies manufacture commercial versions of this or similar dosimeters. Extended rem-counters are both heavier and more expensive than standard instruments.

3.4.2.2.2 Bonner Sphere Spectrometer

A Bonner sphere spectrometer is a few-channel neutron spectrometer, consisting of up to 15 spherical moderator spheres made from PE, some with metallic inserts to extend the energy range beyond an energy of a few MeV (Fig. 3.10).

Fig. 3.10
figure 10

A Bonner-sphere spectrometer consisting of 10 different spheres with amplifiers and data acquisition. [18]

Full size image

Each sphere has a characteristic, energy dependent response curve for neutrons (Fig. 3.11). The count rates for a thermal detector in each of the spheres can be unfolded (a mathematical inversion procedure) to yield an estimate of the energy-dependent neutron fluence spectrum. The dose equivalent can then be determined from the spectrum by multiplying it channel by channel with the corresponding fluence-to-dose equivalent conversion coefficient [25]. The application of a Bonner sphere spectrometer to neutrons from accelerators has been described in [26]. This procedure is time consuming and needs the additional step of spectrum unfolding. It is only employed in research projects when a good knowledge of the spectrum is necessary.

Fig. 3.11
figure 11

Energy-dependent response functions of spheres constituting a Bonner-sphere neutron spectrometer with 12 different spheres. The diameter of the spheres is given in inch. (1 in = 2.54 cm). (From [25]. Figure reproduced with permission from Elsevier Science & Technology Journals)

Full size image
3.4.2.2.3 Proton Recoil-Based Neutron Detectors

In collisions, energetic neutrons can transfer a large part of their kinetic energy to protons. The recoil proton deposits the kinetic energy by charged particle interactions. Building on this principle, some neutron dosimeters for energies of more than 100 keV consist of a neutron-to-proton converter made from a hydrogen-rich material, combined with a charged particle detector.

An active personal dosimeter combining Si diodes with 6Li converter for thermal neutrons, PE converter for fast neutrons and uncovered diodes for photon detection is described in [27, 28].

As an alternative to a Bonner sphere neutron spectrometer, the proton recoil spectrometer can be employed for neutron energies of more than one MeV [23]. They consist of a proportional counter in which the pulse-height, proportional to the energy of the recoil proton, is registered. After unfolding, the pulse height spectrum yields the energy-dependent neutron fluence spectrum. The advantage over a Bonner sphere spectrometer is, that the pulse height spectrum has more channels, which makes the unfolding process more robust and results in a better energy resolution. In a radiation field where low-energy neutrons make an important contribution to dose equivalent the proton recoil spectrometer can be combined with a few Bonner spheres for energies up to the MeV-range.

The functions of recoil and detection are combined in an ionisation chamber filled with hydrogen gas. This chamber measures the absorbed energy of neutrons by registering their recoil protons released from hydrogen molecules. It is also sensitive to photons and electrons and must be calibrated with a field-specific calibration coefficient. This technique has been employed to measure dose equivalent in radiation fields generated by particle accelerators [24].

3.5 Radiation Protection at Accelerators

At high doses, the exposure to ionising radiation may lead to tissue effects, at low doses and dose rates, stochastic health effects are possible and become more probable with increasing effective dose received. To protect accelerator personnel and members of the public, living in the vicinity of the accelerator, from the detrimental effects of ionising radiation, multiple radiation protection strategies are employed:

First, one strides to reduce one of the main sources of ionising radiation at an accelerator, beam loss. Where this is not possible for physical or operational reasons, beam loss is concentrated at a few locations with the help of collimators (Sect. 2.6.1). These are surrounded with shielding to absorb the generated particle cascades.

The remaining radiation environment at an accelerator consists of radiation penetrating the shielding or released from the accelerator area by way of activated air or cooling water. The largest hazard activated material because many workers are exposed to its decay radiation during repair and maintenance of accelerator components.

3.5.1 Shielding Against Prompt Radiation

In contemporary particle accelerators with high beam energy and intensity, the level of ionising radiation from beam loss during accelerator operation is generally severe enough to present a danger for health and life. Consequently, accelerators are installed either underground, or behind shielding walls from iron, concrete and earth, to dilute and absorb the secondary particle cascades following beam-matter interactions.

Methods and results for the design of accelerator radiation shielding fill numerous scientific publications, summarised in [3, 7, 10, 31, 34]. An estimate of the required shielding can be obtained by estimating the flux of secondary particles emerging from a primary beam collision, to determine the attenuation of this flux by shielding material and to convert it into dose equivalent outside of the shielding. Radiation attenuation models have the general form

$$ H\left(r,\theta \right)=\frac{1}{r^2}{H}_0\left({E}_p\right)g\left(\theta \right)\exp \left(-\frac{d}{\lambda\ \sin \theta}\right) $$

Here, H is the expected dose equivalent rate outside of a shielding with a thickness of d under the emission angle θ from the beam loss point or target. H0 is the energy-dependent emission of secondary particles from the target, conveniently expressed as dose equivalent, and g(θ) is the angular dependence of the emission. The product H0 g(θ) is called the “source term”. λ stands for the attenuation length of radiation in the shielding material. The parameters in the equation depend on the type of accelerated particle because the secondary particle spectra differ strongly between electron-, proton- and heavy-ion accelerators. In Table 3.2, a widely used set of parameters for the shielding of proton accelerators is reproduced from [7].

Table 3.2 Parameters of a frequently used shielding formulas for proton accelerators
Full size table

From Table 3.2 one can see that the attenuation length λ =50 cm for concrete, attenuating a hadronic cascade laterally to the beam direction to 1/e (36%).

Approximate shielding models deliver order-of-magnitude estimates of dose equivalent rates, from which the required thickness of simple concrete and iron shielding walls can be estimated within an uncertainty of one attenuation length. Their simplicity allows to evaluate them with a pocket calculator or a spreadsheet, but they cannot cover more complex accelerator layouts, with openings, access labyrinths, connected galleries and distributed locations of beam loss. For complex shielding arrangements one must resort to Monte-Carlo radiation transport programs such as FLUKA [31, 35] or MCNP [32]. The basic idea behind these programs is to describe the development of the secondary particle cascade, by following the “history” of test particles. At every interaction step, the numerical values of the energy and angle-dependent interaction cross sections are drawn randomly from the respective probability distribution functions. Many thousands or millions of particle histories are followed to yield average results of particle fluence, absorbed dose and activation. By multiplying the estimated fluence with conversion coefficients, one can derive operational radiation protection quantities like dose equivalent. An example of the result of a Monte-Carlo transport calculation for radiation shielding is seen in Figure 2.16.

These simulation programs have been developed over decades by collaborations of scientists and they are regularly validated by comparison with actual dose measurements at existing facilities. Their scope goes beyond the calculation of dose equivalent outside of shielding structures, they also allow the estimation of material damage by accelerator beams, or the dose equivalent from activated materials.

3.5.2 Protection Against Ionising Radiation from Activation

In all types of accelerators, low-energy radiation is emitted in form of photons and electrons by activated accelerator components and structural material. Personnel enters in contact with these radiation fields when accelerator components need maintenance, repair, or have to be exchanged during planned or unplanned shutdowns of the accelerator. In these periods, accelerator personnel accumulate the major part of annual personal doses.

The methods of radiation protection against low-energy radiation are similar to those applied in nuclear facilities, or medical environments: One tries use distance to the source, radiation shielding and limitation of the exposure time to keep the personal doses as low as reasonable achievable (ALARA).

  • Establishing sufficient distance to the source of radiation is nearly impossible if the personnel must intervene on activated accelerator material. A possibility for optimisation is the use of teleguided manipulators or robots to accomplish some of the mechanical work. The financial cost and the required additional time make robotic solutions interesting only in those cases where very high dose rates are prevalent, for example on targets and beam dumps.

  • Where possible, shielding against photon radiation is employed, in form of lead-equivalent blankets or mobile walls made from heavy metals. This proves often as unpractical, especially when the source of radiation is widely distributed, for example activated, large accelerator components or the walls of the accelerator tunnel.

  • Given the difficulties to apply distance and shielding, the most effective method for the optimisation of personal dose during particle accelerator maintenance is usually the reduction of exposure time. This can be achieved by well-trained personnel, who may have previously rehearsed the operations in the radiation field on non-radioactive mock-ups. Many nuclear establishments have made the experience that the rigorous preparation of work in a radiation field benefits the reliability of the operation, and that the time and money spent for the minimisation of the radiation dose is paid back by an increased efficiency of the intervention.

3.5.3 Control of Radioactive Material

Owners of radioactive material are obliged to exercise close control over the material, to prevent its loss or dissemination in the public. Activated material falls under these regulations as soon as its activity exceeds a legal threshold, the exemption level. If the activity or activity concentration for one isotope exceeds the corresponding exemption level, the material is considered as radioactive. For the European Union, the exemption levels are published in the EURATOM Basic Safety Standards [30]. Radioactive material cannot be traded freely (for example as scrap metal or as filler material in constructions, but it must be stored in intermediate or final radioactive waste depositories. These are operated by organisms authorised by national authorities, and the storage of the material must be paid for.

The transport of radioactive material is also strictly regulated. In Europe, an agreement for the road transport of dangerous goods was concluded in 1957. Its original title in French reads “Accord européen relatif au transport international des marchandises Dangereuses par Route “and it is best known under its acronym “ADR” [29]. Radioactive materials fall in the class 7 of dangerous goods. Depending on the activity concentration, total activity, and dose equivalent rate on the outside of the packaging, shipping of radioactive material must meet several requirements. In the worst case, it can only be transported in special containers by authorised companies.

The price to be paid for the storage and transport of radioactive material make that minimisation of its quantity and its movements is not only a safety consideration, but may also have a positive financial effect.