Radiation Protection at Petawatt Laser-Driven Accelerator Facilities: The ELI Beamlines Case

Abstract

ELI Beamlines is a newly constructed petawatt (PW) laser-based accelerator facility. Its flagship laser has a nominal peak power of 10 PW with a pulse duration of 150 fs. The radiation fields emerging from the laser-target interactions will be pulsed and made of mixed particles of high intensity and energy, thus posing new and daunting challenges compared to conventional radiation protection. The civil engineering constraints of a laser facility bring localized weaknesses to the shielding structures. The ultrashort pulses in conjunction with ultrahigh dose rates strain the capabilities of the commercially available radiation detectors. The complexity of modeling laser-target interactions and the resulting ionizing radiation complicates the radiation protection calculations. This paper is a review of the radiation protection challenges at ELI Beamlines, the adopted solutions, and the practices to minimize the harmful effects of ionizing radiation linked to its activities. Some of the presented solutions may have universal validity, while others are specific to this site and can only serve as an inspiration.

Keywords:

• Radiation protection
• laser-driven accelerator
• ultrashort pulsed radiation
• petawatt lasers
• ELI Beamlines

I. INTRODUCTION

In 1998, Backus et al.¹ published the first review of ultrahigh-power lasers, which identified only one petawatt (PW)-class laser in operation worldwide. Less than 3 decades later, PW lasers are numerous, and projects for future laser facilities are regularly being proposed.^2,3 Both research and industry, in fact, are pushing toward the development of evermore powerful lasers with higher repetition rates and higher intensities. These remarkable facilities enable pioneering research and technological advances in a variety of fields, including but not limited to, plasma and fusion research, atomic molecular and optical physics, femtosecond chemistry, astrophysics, high-energy physics, materials science, biology, and medicine.⁴

In fact, ultrahigh-power lasers are exploited to produce and accelerate particles, thus constituting a novel family of accelerators. Alongside the rapid improvements in laser-plasma acceleration technology, there is a rapid increase in the correlated radiological risks. In fact, the expected high energies (up to tens of gigaelectron-volts for electrons and on the order of 1 GeV for proton beams) and current repetition rates (the estimated number of particles per laser shot can be 10⁹ to 10¹⁰ and 10¹⁰ to 10¹² for electron and proton beams, respectively) require accurate radiation protection (RP) studies to protect personnel, the general public, the environment, and equipment, all the time assuring efficient and smooth operations.^4–8

This paper provides a comprehensive overview of the RP hazards at the Extreme Light Infrastructure (ELI) Beamlines laser-driven accelerator facility. Attention is given to the adopted RP measures, procedures, equipment, and tools used to mitigate these hazards.

II. ELI EUROPEAN RESEARCH INFRASTRUCTURE CONSORTIUM

The design and construction of the ELI facilities started in the early 2000s, but it was only in 2021 that the European Commission established the ELI project⁴ as a European Research Infrastructure Consortium⁹ (ERIC), thus creating the ELI ERIC. The ELI ERIC is the largest international civilian laser-based user facility network. The establishment of the ELI ERIC enables researchers and industries to access the world’s largest collection of high-power and ultrafast lasers. As an international organization, its member countries contribute scientifically and financially to the consortium. The founding members are the Czech Republic, Hungary, Italy, and Lithuania, while Germany and Bulgaria are founding observers with the aim to fully join at a later date. Other European countries have expressed interest in joining as well.

The ELI ERIC facilities are located in Hungary, Romania, and the Czech Republic. The Hungarian laser facility, ELI Attosecond Light Pulse Sources, uses pulses of attosecond length for material sciences and biology research.¹⁰ ELI Nuclear Physics in Romania focuses on photonuclear physics, bringing together high-power lasers and nuclear physics.¹¹ Located in the Czech Republic, ELI Beamlines¹² aims to investigate high-field high-density physics, developing high-brightness sources of X-rays, as well as proton, electron, and ion beams.

III. ELI BEAMLINES

ELI Beamlines¹² is a laser-driven user facility located south of the city of Prague, Czech Republic. Its research activities can be classified in complementary areas of scientific interest: development and testing of novel technologies for multi-PW laser systems, plasma physics, high-field physics experiments, production of femtosecond secondary sources of ionizing radiation (extreme ultraviolet radiation, X rays, gammas, electrons, and protons) to be used in interdisciplinary applications in physics, biology, medicine, and material sciences.¹³ In-house experiments started as early as the first half of 2018. The first user calls were issued in 2019, and the first user experiments were successfully performed later that same year. Licenses to operate the stations producing ionizing radiation were secured from the Czech State Office for Nuclear Safety (see Sec. IV).

ELI Beamlines houses four main laser systems (see Sec. III.A) that serve a dozen experimental stations (see Sec. III.B) with the possibility of adding more in the future. The laser systems and the experimental stations are located in a dedicated building along with the necessary auxiliary systems, such as cooling, ventilation, and vacuum systems (see Sec. VII).

III.A. Laser Systems

The current and nominal parameters of the four main laser systems are listed in Table I. L1-Allegra, simply meaning cheerful in Italian, was developed in-house by the ELI Beamlines laser team and has been in operation since 2018 (Ref. 14). It is designed to generate <20-fs pulses with energies exceeding 100 mJ per pulse at a 1-kHz repetition rate. The L2-DUHA laser (Dual-beam Ultra-fast High-energy OPCPA Amplifier, “duha” meaning rainbow in Czech) was designed inhouse to provide 100-TW-level pulses at a 50-Hz repetition rate and an 820-nm wavelength.¹⁵ The L2-DUHA is currently being installed, with the start of the commissioning in 2023. The L3-HAPLS (High-Repetition-Rate Advanced Petawatt Laser System) is designed to deliver PW pulses with a peak energy of 30 J and <30-fs pulse lengths at a repetition rate of 10 Hz (Ref. 16). This system was developed at the Lawrence Livermore National Laboratory in cooperation with ELI Beamlines. The L3-HAPLS has been in operation since 2019. Finally, the L4-Aton, named after the Egyptian sun god, is in trial operation. It will eventually deliver three different laser beams. The target peak energy of the most powerful one is an impressive 10 PW with a pulse duration of 150 fs (Ref. 17).

TABLE I Present and Target Parameters of the Laser Systems at ELI Beamlines

Laser	Peak Energy (J)	Peak Power (TW)	Maximum Repetition Rate (Hz)
L1-Allegra
(Present)	0.03	1.5	10^3
(Target)	0.1	5	10^3
L2-DUHA (target)	2	10^2	50
L3-HAPLS
(Present)	30	333	3.3
(Target)	30	10^3	10
L4-Aton (target)	1.5∙10^3	10^4	1 shot per minute

III.B. Experimental Stations

Among all the experimental stations at ELI Beamlines, at least nine are expected to generate ionizing radiation. These are summarized in Table II. Photon, electron, proton, and ion beams will be produced and accelerated. Neutrons will be produced as well, but only as secondary byproducts.

TABLE II Experimental Station at ELI Beamlines Expected to Produce Ionizing Radiation

Station	Laser	Main Goal	Reference
HHG	L1-Allegra	Ultrashort pulses of tunable coherent extreme ultraviolet soft X-ray radiation	Ref. 18
PXS	L1-Allegra	High-brightness X-ray beams with energies varying from 3 to about 77 keV	Ref. 19
E2 X-Ray source	L3-HAPLS	Ultrafast and bright hard X-ray	Ref. 20
LUIS	L2-DUHA L3-HAPLS L4-Aton	Laser-driven electron beams for the generation of initially spontaneous, and subsequently, coherent photon radiation	Ref. 21
ELBA	L2-DUHA L3-HAPLS L4-Aton	Acceleration of electrons up to energies of tens of gigaelectron-volts	Ref. 22
ALFA	L1-Allegra	Electron beams up to energies of 50 MeV	Ref. 23
TERESA	L3-HAPLS	Provide proton beams of 10 to 15 MeV and electron beams of 100 to 150 MeV	Ref. 24
ELIMAIA	L3-HAPLS L4-Aton	Protons acceleration up to 60 MeV (phase 1) and up to 250 MeV (phase 2)	Ref. 25
P3	L3-HAPLS L4-Aton	High-field laser-plasma interaction high-energy density physics	Ref. 26

The High-order Harmonic Generation¹⁸ (HHG) experiment is dedicated to the production of ultrashort pulses of tunable coherent extreme ultraviolet soft X-ray radiation using the L1-Allegra. The Plasma X-ray Source¹⁹ (PXS) experiment produces high-brightness X-ray beams with energies varying from 3 to about 77 keV using the L1-Allegra. The E2 X-ray Source, instead, is dedicated to the generation of ultrafast and bright hard X-rays using a PW-class laser with a 10-Hz repetition rate.²⁰ It relies on the L3-HAPLS laser focused on gas targets. The Laser Undulator Incoherent Source (LUIS) experimental station aims to produce laser-driven electron beams for the generation of initially spontaneous, and subsequently coherent, photon radiation.²¹ The electron beams will have maximum energies of 600 MeV with a peak energy spread of less than 1%. The Electron Beam Accelerator (ELBA) for fundamental science and applications will accelerate electrons up to energies of tens of gigaelectron-volts.²² Both the LUIS and ELBA use the L3-HAPLS and in the future will also use the L2-DUHA and L4-Aton lasers.

The Allegra Laser For Acceleration (ALFA) experimental station, instead, uses the L1-ALLEGRA laser to generate a 50-MeV electron beam.²³ The Testbed for high Repetition-rate Sources of Accelerated particles (TERESA) is designed for proton and electron acceleration²⁴ using the L3-HAPLS laser. Its goal is to provide proton beams of 10 to 15 MeV and electron beams of 100 to 150 MeV. The ELI Multidisciplinary Applications of laser-Ion Acceleration (ELIMAIA), instead, aims at ion acceleration using the L3-HAPLS and L4-ATON lasers; in its first phase it will accelerate protons up to 60 MeV and later up to 250 MeV (Ref. 25). Finally, the Plasma Physics Platform (P3) is a multifunctional experimental infrastructure. With its 50-m³ vacuum chamber, it pursues an ambitious experimental program in high-field laser-plasma interaction as well as high-energy density physics.²⁶ P3 is served by the L3-HAPLS and L4-Aton laser systems and is designed to simultaneously focus up to five different laser beams.

IV. LEGAL REQUIREMENTS AND INTERNAL RP CRITERIA

In the Czech Republic, the State Office for Nuclear Safety (SÚJB, Státní Úřad pro Jadernou Bezpečnost) is responsible for countrywide monitoring, assessment, and control in the fields of personal and environmental protection against the adverse effects of ionizing radiation. Its work is regulated by the Atomic Act 263/2016. on the Peaceful Use of Nuclear Energy and Ionizing Radiation and 20 additional decrees, which comprise the Nuclear Law.²⁷ One of the most relevant documents to ELI Beamlines is Decree 422/2016 Coll. on the RP and security of a radioactive source. The Nuclear Law complies with the relevant recommendations and legislation from the International Commission on Radiological Protection, European Atomic Energy Community, International Atomic Energy Agency, and the European Union.^28,29

The RP group at ELI Beamlines works in close collaboration with SÚJB, which carries out regular inspections and audits and grants the commissioning and operational licenses for each experimental station. The details of the implementation of individual commissioning phases and any nonroutine activities are covered by internally issued temporary work authorizations, which include a justification of the activity to be carried out, a full risk analysis, and the implemented safety measures. The laser building (see Sec. VII), which houses the laser systems and the experimental stations, is divided into supervised and controlled areas. The delineation of the areas follows the law guidelines. Supervised areas are established wherever the effective dose could exceed 1 mSv/year or the equivalent dose to eye lens, skin, or extremities could exceed 1/10 of the legal limits (see second column in Table II). Instead, controlled areas are delineated where the effective dose could exceed 6 mSv/year or the equivalent dose could exceed 3/10 of the limits for skin or extremities or 15 mSv for eye lens (see second column in Table II).

The second column of Table III lists the maximum allowed legal exposure limits for radiation workers in the Czech Republic. However, in agreement with the national authorities and always following the “as low as reasonably achievable” (ALARA) principle,³⁰ ELI Beamlines aims at the more stringent internal RP criteria listed in the third column of Table III. This optimization of protection and safety is achieved by a careful construction of the experimental halls and control rooms, an attentive choice of shielding materials (see Sec. VII), and detailed work plans for each experiment where worst-case scenarios are considered. During upgrades, commissioning, and even unscheduled interventions, risk assessments are used to minimize the duration of work in radiation areas, allow for adequate cooldown time depending on the urgency of the intervention, and optimize the use of necessary human resources.

TABLE III Maximum Legal Exposure Limits for Radiation Workers in the Czech Republic²⁷ and ELI Beamlines Internal Limits*

	Czech Legislation	ELI Beamlines Internal Limits
Effective dose (external and internal)	20 mSv/year (or 100 mSv per five consecutive years with maximum of 50 mSv/year)	1 mSv/year
Equivalent dose to eye lens	100 mSv/5 consecutive years and maximum 50 mSv/year	15 mSv /year
Equivalent dose to skin	500 mSv/cm^2/year	50 mSv/cm^2/year
Equivalent dose to extremities	500 mSv/year	50 mSv/year

*See Ref. 27 for maximum legal exposure limits for radiation workers in the Czech Republic.

Any exposure of radiation to workers (even if well within legal limits) has to be justified in relation to the urgency and difficulty of the work and the specific expertise of the personnel involved. Additionally, for pregnant or breastfeeding radiation workers, it is forbidden to perform tasks involving possible contamination risks (the legal limit in the Czech Republic for pregnant radiation workers is 1 mSv from when they notify their employer until the end of their pregnancy). Finally, the total maximum effective dose to nonradiation workers and the general population is less than 0.1 mSv/year (the legal limit in the Czech Republic is 1 mSv/year for all sources of artificial radiation not connected to medical procedures).

The head of the RP group is the appointed Radiation Safety Officer, responsible for the safe use of radiation and radioactive materials as well as regulatory compliance. An internal Radiation Committee has been established to advise facility management on radioprotection matters, recommend policies, review accident investigations, and evaluate potential protective measures. The members, appointed by the institute director, are the main RP stakeholders, such as representatives of the experimental teams and the RP group.

V. RP CHALLENGES AT ELI BEAMLINES

When designing and implementing the RP and safety measures at ELI Beamlines, existing international recommendations and recognized best practices were followed. However, high-power and high-intensity laser facilities are new and distinctive radiation installations, for which recognized standards are currently being drafted. Therefore, it was necessary to develop ad hoc solutions for the unique RP challenges faced at ELI Beamlines. An overview of these challenges and the adopted solutions are presented.

V.A. Wall, Floor, and Ceiling Penetrations

When designing the civil structure of the facility, RP concerns were carefully considered and attention was paid to shielding requirements. Nevertheless, hundreds of penetrations are present in the ceilings and walls. While these are needed to transport the laser beams and auxiliary systems throughout the building, at the same time they introduce significant weaknesses in the shielding structures. Further, most of the penetrations are wide, typically measuring 1 × 1 m, with the largest ones reaching 9 m². These apertures are effectively empty since they accommodate vacuum tubes, typically dedicated to laser beam transport. Consequently, many are at laser beam height (1.3 m from the floor), where the highest dose rates are expected. Besides, these penetrations must be built straight, since each turn of the laser is not only costly, but it significantly deteriorates the properties of the laser beam.

Further penetrations are used for the cooling system, transport of gases needed in the experiments, and for the many air-conditioning units needed to comply with the EN ISO 14644 standard for clean rooms³¹ (see Sec. V.A). Also, narrow but nonnegligible channels run beneath the experimental floors for cable passageways. All these wall and floor weaknesses offer paths for the radiation to leak from one area to another. Each penetration was studied separately considering hazards related to ionizing radiation, electromagnetic pulses (EMPs), and fire. Different solutions were adopted depending on the dimensions and positions of the penetrations and on the equipment within. Concrete blocks have been placed in the cavities and local shielding has been designed wherever possible. Polyethylene and borated polyethylene for neutron shielding have been used to fill some of the ceiling penetrations (loose material with a density of 430 g/l that can be tightly packed in the gaps). Finally, some penetrations, such as the ventilation ducts, cannot be closed off nor shielded. Active monitoring detectors have been placed in strategic locations close to these penetrations, where the dose rate is expected to be higher and/or where personnel may be present.

Figure 1 shows the simulated ambient dose equivalent [H*(10)] rates in and around the LUIS experimental hall. Simulations were performed using the FLUKA Monte Carlo code^32,33 (see Sec. VI). A long channel for signal and power cables runs under the floor along the direction of the beamline into the control room. The channel is 60 cm wide and 20 cm deep (Fig. 3a). Radiation may leak through the channel into the control room. The increase in the dose rate in the control rooms was estimated to be on the order of 10 nSv/pulse. For this reason, a concrete shielding wall was designed. It measures 80 × 60 × 200 cm, and its location is shown in Fig. 2. The estimated reduction in H*(10) in the control room is of about two orders of magnitude. Finally, as an extra and redundant shielding measure, concrete blocks were placed inside the channel (Fig. 3b) in compliance with the ALARA principle (see Sec. IV).

Fig. 1. Monte Carlo estimation of the H*(10) rate in the LUIS experimental hall. An electron beam with a peak energy of 400 MeV with 100 pC per pulse was used as the source term for this study. No additional shielding is installed between LUIS and the control room. Radiation leaks through the floor penetrations toward the control room.

Fig. 2. Monte Carlo estimation of the H*(10) rate in the LUIS experimental hall. An electron beam with a peak energy of 400 MeV with 100 pC per pulse was used as the source term for this study. A concrete shielding wall is shown in the figure. The H*(10) rate reduction in the control room is of two orders of magnitude (compared with Fig. 1).

Fig. 3. Picture of the floor penetration in the LUIS experimental hall (a) without and (b) with concrete blocks. The same set is installed on the control room side of the penetrations, but inverted, thus forcing juggled routing of the cabling. These concrete blocks are a redundant shielding easily added in compliance with the ALARA principle (see Sec. IV).

V.B. Source Terms

The ionizing radiation generated in laser-target interactions has a few important characteristics. First, it is generated in very short pulses.³⁴ Second, it comprises several types of particles (e.g., photons, electrons, and neutrons) spanning over a very large energy range.³⁴ Since the beginning, RP assessments at ELI Beamlines have been based on Monte Carlo simulations. The descriptions of ionizing radiation used as inputs to the simulations (called source terms) were either extrapolated from previous experiments or estimated via particle-in-cell simulations,³⁵ which are used to model the interaction of lasers with targets. In fact, these interactions are themselves the subject of research.³⁶ Hence, they suffer from limited experimental reproducibility and their modeling faces mathematical and computational challenges.³⁷

As a result, the source terms used for RP assessments suffer from intrinsic uncertainties much larger than those affecting source terms used in conventional accelerators. Therefore, a conservative worst-case scenario approach was used for the source energy, charge, repetition rate, and geometrical beam characteristics. Figures 4a and 4b show how minimum changes to the source term can have a ripple effect on the predicted doses. The estimated H*(10) rate outside the exit window of the interaction chamber of ALFA is plotted considering an electron beam with 3.5-MeV average energy and 2.5-MeV full-width at half maximum, but with different beam divergences (12 and 100 mrad). The exact beam divergence is unknown and may vary significantly between data-taking runs. As seen in the plots, the variations in the expected radiation field are remarkable.

Fig. 4. Monte Carlo ambient dose equivalent rate outside the exit window of the ALFA interaction chamber assuming an electron beam with 3.5-MeV average energy and 2.5-MeV full-width at half maximum. The beam divergence was set to (a) 12 mrad and (b) 100 mrad.

V.C. Ultrashort Laser Pulses

Multi-TW and PW lasers with ultrashort pulse durations (tens or hundreds of femtoseconds) interacting with solid, liquid, and gaseous targets will produce high-intensity, high-energy, and pulsed radiation fields, which pose new challenges compared to conventional RP.

The interaction of subpicosecond laser pulses with solid targets generates strong EMPs that can cause damage to electronic devices.^38–40 This imposes severe constraints on the detectors that can be employed in a high-power laser facility. Until now, RP detectors have not been tested for resistance to EMPs, as these were never associated with the generation of ionizing radiation. Different mitigation strategies could be pursued.³⁸ The most relevant for RP purposes, is to shield active detectors within Faraday cages to try to protect them from the EMPs. Another approach is to use passive detectors, whenever possible, as these have never been shown to suffer from EMPs. Both approaches are used at ELI Beamlines. Additionally, all the experimental halls are built as standalone Faraday cages to protect both personnel and equipment, including RP detectors for area monitoring.

The ultrashort laser pulse duration implies that the ionizing radiation is also produced in very short pulses. Real-time radiation detection instruments available on the market are typically designed and optimized for continuous fields. The International Organization for Standards (ISO) defines a continuous radiation as ionizing radiation with a constant dose rate at a given position in space for a time interval longer than 10s (Ref. 41). The main challenge for dosimetric systems is to process high detector loads within short time periods (ultrahigh dose rates), maintain an appropriate dead-time behavior, ensure pile up suppression, and reliably distinguish signal from noise.⁴² Many of the commercially available electronic dosimeters can detect gradual intensity changes, but do not provide reliable dose values in these pulsed fields.⁴³ Therefore, it is good practice to use passive systems in conjunction with active ones (see Sec. X for further details). Concurrently, the RP group at ELI Beamlines is participating in international efforts to develop new metrological tools needed for traceable absorbed dose measurements^44,45 and to test existing and newly developed detectors at ultrahigh pulse dose rate beam facilities.

V.D. Radiation Damage to Electronics

The potential danger posed by ionizing radiation to electronics was first postulated in 1962 (Ref. 46) and has moved center stage in large scientific laboratories with the CERN Neutrinos to Gran Sasso (CNGS) accident at CERN in 2007 (Ref. 47). Therefore, the risks of damage to electronics, both due to single-event effects (SEEs) and cumulative effects⁴⁸ is regularly investigated at ELI Beamlines in order to identify situations that could potentially harm the regular operations and to implement mitigation actions. This risk is another argument in favor of using a combination of active and passive detectors for area and personal monitoring. These studies are important also for the experimental teams to understand, limit, and possibly avoid radiation-to-electronics damage.

High-energy (>20 MeV) Hadron (HEH) fluence is used to characterize stochastic SEE failures. Figure 5 shows the estimate of the HEH fluence generated in a solid target experiment with the L4-ATON in the E3 hall. The values are given in the number of particles per square centimeter per laser shot. Values below 10⁷ particles/cm²/year are considered acceptable for nonradiation-hard electronics.^49,50 Therefore, the number of shots per year limits the possible placements of the radiation detectors, and vice versa, the need to monitor the radiation level in certain areas, while avoiding failures in the monitoring devices that may limit the run time.

Fig. 5. HEH fluence in the E3 experimental hall for a solid target experiment driven by the L4-ATON laser, estimated via FLUKA simulations.

Total ionizing dose and 1-MeV-neutron equivalent fluence in silicon (Si-1 MeV n_eq), instead are used to address the cumulative effects. Assuming 36 000 laser shots of the L4-AATON in a year (2 h/day and 300 days/year at one shot per minute), the estimated Si-1 MeV n_eq fluence is shown in Fig. 6. Areas with values of 10¹⁰ particles/cm²/year are considered safe for commercial electronics.

Fig. 6. Si-1 MeV n_eq fluence in the E3 experimental hall for a solid target experiment driven by the L4-ATON laser, estimated via FLUKA simulations.

The need to develop a deeper understanding of the radiation damage to electronics has led the scientific community to launch the RADNEXT project.⁵¹ The RP group at ELI Beamlines is an active participant in this H2020 INFRAIA-02-2020 infrastructure, which is aimed at building an international community of laboratories, universities, and industries working on the radiation-to-electronics-damage problem.

V.E. Clean Rooms

High-power lasers are transported in vacuum and need to be operated in clean room conditions in order to not deteriorate the laser quality and not damage any device. All experimental halls at ELI Beamlines are classified as ISO 5 or ISO 7 according to the ISO standard for clean rooms.³¹ Consequently, shielding materials and any monitoring equipment must comply with the cleanliness requirements. They must have smooth surfaces for easy cleaning, they should be resistant to cleaning chemicals, and they should release neither dust nor particles into the air. Effectively, for the construction of shielding and beam dumps, this implies that concrete blocks can be used only if appropriately coated and/or covered. Granite, only slightly more expensive than concrete but with approximately the same shielding properties, is considered a suitable alternative. Adequate surface finishing allows for easy polishing while demonstrating acceptable chemical absorbency. Besides, it is widely available and can be machined to the needed shape and dimensions. For what concerns monitoring devices, all the passive area dosimeters are packaged in clean aluminum or plastic containers. The active detectors have been carefully selected based on manufacturing specifications and were thoroughly cleaned before installation.

The requirement for cleanliness is even more pronounced for objects inside the interaction chambers. Most materials are not compatible with the vacuum requirements, including most plastics, due to their outgassing. Thus, further constraints are imposed on the choice of materials, e.g.,, on the cartridges of passive detection systems.

Finally, clean room attire and laser-safety goggles limit movement and visibility, extending the time needed for any intervention (see Fig. 7). This limitation must be factored in when planning activities in radiation areas. On the other hand, the compulsory protective clothing (including gloves) reduces the risk of skin contamination when handling activated material.

Fig. 7. Operators at work on the laser system. Clean room ISO 7 attire and laser-safety goggles limit movement and visibility, increasing the time needed for each intervention, and consequently, the exposure time.

VI. MONTE CARLO SIMULATIONS

ELI Beamlines is a new facility still in the commissioning phase. At the time of writing, only two of the main laser beamlines were fully operational and were still being expanded and upgraded to reach their nominal performance, while experimental stations were in various phases of commissioning. Therefore, RP assessments have been mostly driven by Monte Carlo simulations. The RP group at ELI Beamlines uses mainly FLUKA (Refs. 32 and 33). FLUKA is a general-purpose Monte Carlo code for the interaction and transport of hadrons, heavy ions, and electromagnetic particles from a few kilo-electron-volts to cosmic-ray energies in arbitrary materials. The code is used worldwide for RP calculations at accelerator facilities. In fact, it is one of the few codes that provides combined prompt and residual dose rate calculations as well as the simulation of particle interactions in the energy range needed. FLUKA is able to transport 60 different elementary particles and any heavy ions.⁵² Also, it includes a solver routine for the Bateman equations,^32,33 estimators relevant for radiation damage to the electronics, and a user-friendly graphical interface FLAIR (Ref. 53).

A single, highly modular, and highly scalable FLUKA simulation input file is used for all RP studies at ELI Beamlines. In more than 50 000 lines of code, all experimental halls and stations have been modeled and relevant physics settings activated. This ensures a single and uniform environment across all performed RP studies where a sufficiently detailed and realistic simulation model of the laser building is used. The FLUKA geometry model of the experimental halls located in the basement of the laser building is shown in Fig. 8.

Fig. 8. FLUKA geometry model of the experimental halls at ELI Beamlines. This corresponds to the basement in Fig. 9.

Several tasks are performed by means of Monte Carlo simulations. One of the most relevant is the characterization of radiation field maps based on source terms agreed upon with the experimental teams and relying on either extrapolation from previous experiments with similar conditions or particle-in-cell simulations. Radiation field maps are generated for several quantities, like H*(10), HEH fluence, and Si-1 MeV n_eq fluence, and are then used for the assessment of the risk for humans and machines. The identification and the design of possible mitigation options are also based on Monte Carlo simulations. This includes identifying more suitable locations for devices and the design of radiation shielding. Additionally, FLUKA simulations are used to estimate the possible induced radioactivity of materials exposed to the radiation generated in laser-target interactions and advanced planning for the management of the activated materials. In fact, FLUKA offers unique capabilities for activation studies as the prompt and residual radiation particle cascades are simulated in parallel and are based on microscopic models for nuclide production and analytical solutions of the Bateman equations for activity buildup and radioactive decay.^32,33

VII. EXPERIMENTAL HALLS: STRUCTURE AND SHIELDING OF THE PROMPT RADIATION

To fulfill the aforementioned RP criteria as well as scientific and technological needs, civil engineering constraints, and national and international regulations, many Monte Carlo simulations and analytical calculations have been performed.^32,33,54 The results of these calculations drove the design and the construction of the laser building. All calculations were performed under conservative assumptions, thus ensuring comfortable safety factors. Maximal energy and intensity values were used for all source terms to obtain a conservative estimate of the radiological hazards. Laser operations were assumed to cover up to 250 days/year and 8 h/days, if not specified differently by the experimental teams. The soil surrounding the basement provides extra shielding but was not taken into account during the design of the facility. Finally, all penetrations (see Sec. VI) were considered empty, i.e., without shielding material.

Figure 9 shows a schematic view of the laser building. The building is made of reinforced ordinary concrete. During construction, special attention was paid to the concrete density (at least 2430 kg/m³ for fresh concrete, corresponding to 2320 kg/m³ for mature concrete) and to the homogeneity (uniform compaction) of the built walls. Poured concrete was used, except for the areas requiring heavier shielding, where high-density (4000 kg/m³) concrete was used. In this case, to prevent the settling of the aggregates and thus compromise the shielding properties, prefabricated blocks of material were used.

Fig. 9. Diagram of the Laser building, indicating the locations of the main technological units.

The ground floor houses the main laser systems and the experimental stations ALFA and TERESA, and it is classified as a supervised area (see Sec. IV). Instead, all the other experimental stations are in the basement, which is 7.6 m underground and is classified as a controlled area (see Sec. IV). The external and internal walls separating the six main experimental halls are typically 1.60 m thick. The minimal concrete thickness between the floors is 0.5 m. These thicknesses allow for access to any experimental hall regardless of the operational state of the adjacent ones.

Each experimental hall has a dedicated control room. From the control room, access to the hall is through a maze cut by three doors: an EMP door, a neutron shielding door, and a fire safety door. The neutron shielding doors are made from 5 cm of high-density pressed laminated wood with a hydrogen content of 6%, high content of low-Z elements (carbon 50%, oxygen 43%), and a mass density of ~1000 kg/m³, suitable for neutron shielding.^55,56 Each hall is also equipped with large doors allowing for the transport of bulky material. Also, this access is protected by a concrete maze and closed by large EMP shielding doors. These large EMP doors and the ones separating the control rooms from the experimental halls are all connected to the interlock system (see Sec. IX). Where relevant, dismountable shielding walls and beam dumps are used to avoid the activation of walls and floor. Standard-density (2350 kg/m³) concrete blocks and granite are used for this purpose.

The effectiveness of the proposed measures was thoroughly investigated. Figure 10 shows an example of such a study for a hall with a 500-MeV electron source. In the experimental hall, the prompt H*(10) rate is higher than 25 µSv/day (assuming 1800 shots per day). The rate drops to 2.5 to 4.0 µSv/day in the plant room, where the support facility systems (air-conditioning and cooling units, roughing vacuum pumps, etc.) are located. The experimental hall and the most dangerous sections of the plant room are interlocked during run time and access is forbidden (see Sec. IX). The adjacent corridor has an estimated H*(10) rate of less than 1 µSv/day.

Fig. 10. H*(10) rate simulated in an experimental hall hosting a 500-MeV electron source and 1800 shots per day (vertical view). The corridor and control room are treated as a normal workplace, i.e., personnel will be present for daily 8-h shifts.⁵⁴

These studies were done simulating the entire electromagnetic cascade with all secondary particles, including low-energy neutron interactions (<20 MeV). Photons and electrons were transported down to energies of 1 keV (kinetic energy). The electrons, photons, and neutrons are the main contributors to the dose rate. The shielding wall between the hall and the control room is thick enough to absorb the electromagnetic components and the neutrons impacting directly upon it. The electromagnetic component is, therefore, well contained in the experimental hall (which is empty during operations). Instead, through multiple scatterings, the neutron prompt radiation “leaks” through the ventilation ducts above the control room toward the plant room, which is a low-occupancy area (see arrows in Fig. 10). The corridor walls have been designed such that the H*(10) rates in the control room are comparable with the background rate, as desired and as confirmed by the simulation (see Fig. 10). Therefore, the design of the building and the materials used meet RP safety requirements.

VIII. ACTIVATION, CONTAMINATION, AND WASTE MANAGEMENT

FLUKA is used for studies of induced radioactivity and nuclide production, nuclide decay, and transport of residual radiation. In fact, the results of these simulations allow for effective planning of maintenance/upgrade interventions in radiation-controlled environments, optimization of the materials used, and management of storage, treatment, and disposal of radioactive waste. The experimental teams and engineers are provided with general guidelines on the choice of materials, based both on the activation cross sections and FLUKA studies.⁵⁷ Preferably, only materials with low radiological impact should be employed. However, since this is not always feasible, dedicated simulations for specific devices and geometry setups and representative irradiation history have been modeled. In some cases, additional moveable local shielding was designed to enable safe work in the vicinity of hot spots.

Small-size activated equipment (e.g., target holders, optical components, etc.) will be stored in a dedicated storage area until its activity decays below the exemption limits²⁶ and the equipment is ready for reuse. A tracking database has been prepared for this purpose. An estimated time for the release of the item is provided, which is based on the initial dose measurements and expected nuclide inventory. Large items that cannot be dismounted are regularly monitored, and when needed, local shielding is added. Figure 11 shows an activation study of the first quadrupole in ELIMAIA after 1 year of operation and a 1 h cooldown. The ambient dose equivalent rate is significantly reduced when a stainless steel shielding block is added (see Fig. 12). The shielding will be installed during upgrade and maintenance activities.

Fig. 11. Residual dose rate studies of the first quadrupole of ELIMAIA after 1 year of operation and 1 h of cooldown time. H*(10) rates in its proximity are significant. Extra shielding in this area was deemed necessary during maintenance and upgrade periods (see Fig. 12).

Fig. 12. Residual dose rate studies of the first quadrupole of ELIMAIA after 1 year of operation and 1 h of cooldown time with stainless steel shielding. H*(10) rates are significantly reduced with respect to the scenario without the shielding (see Fig. 11).

To understand the long-term activation issues and to assess the decommissioning needs, Monte Carlo activation studies were performed for the bulky components of the experimental setups, including the beam dumps. A conservative approach was adopted for estimating the radionuclide inventory. It was assumed that the beamlines will operate at the maximum power for 8 h/day, 200 days/year for 25 years. Several cooling times were considered (10 min, 1 h, 1 day, 1 week, 1 month, and 1 year), as well as the possible presence of impurities. Based on these Monte Carlo studies, it is expected that short- and long-term activation will be an issue, particularly for steel components, which will contain relative isotopes such as ⁴⁶Sc, ⁴⁸V, ⁵¹Cr, ⁵²Mn, ⁵⁴Mn, ⁵⁵Fe, ⁵⁹F3, ⁵⁵Co, ⁵⁶Co, ⁵⁷Co, ⁵⁸Co, and ¹⁴⁷Pm. On the other hand, the activation of concrete, polyethylene, granite, and aluminum structures is expected to decay under exemption limits in a short time interval. Based on these preliminary estimates, the activity of each beam dump will be approximately 1 GBq at the time of decommissioning. These expectations will be updated on the basis of more accurate simulations (refined geometry and source term descriptions) and will be integrated with repeated measurements of the activated parts.

A standalone and grinding topic is the radioactive contamination inside the interaction chambers originating from target debris. In fact, laser-target interactions are often destructive. Best practices and protocols for decontamination will be followed. However, since the radionuclide inventory of the contamination will be strongly dependent on the target and laser parameters, more dedicated studies are required. As a precaution, a post-experimental state has been introduced (see Sec. IX). The purpose of this mode is twofold: to allow short-lived radionuclides to decay and to give RP experts time to perform the necessary measurements and possible decontamination procedures before access is granted to other personnel.

In compliance with the legal requirements, an official plan for the decommissioning of ELI Beamlines was prepared by the RP group. It was approved by the national authority for radionuclide waste storage and its cost was calculated by an authorized company for reprocessing of radioactive materials. Once the decommissioning is completed, it is expected that it will be possible to use the premises without any restrictions. The decommissioning plan will be regularly reviewed and updated every 5 years.

IX. PERSONAL SAFETY INTERLOCK

The fundamental active layer of protection is the Personal Safety Interlock (PSI) system, which is intended to recognize the presence of hazards to personnel, initiate automatic protective measures, trigger alarms, and eliminate the hazard automatically, if possible. The covered hazards include, but are not limited to, ionizing radiation, laser, vacuum, and high voltage. In fact, the PSI is designed as an all-inclusive, redundant, fail-safe, independent electronic system compliant with an international standard for Functional Safety of Electronic Safety Systems.⁵⁸ Implementation, installation, and independent safety audits of the system have been outsourced to an external company.⁵⁹ However, the ELI Beamlines Safety Team prepared the conceptual design and was fully engaged in the risk assessments and definition of the safety functions.

A hybrid hardware-software-based solution was adopted. From system architecture, logic, and the software side, each experimental hall (with its unique set of hazards) is managed independently, but information from adjacent halls is shared. The system logic defines rules for the halls’ access and initiations of safety functions that prevent the occurrence of hazardous situations. In its simplified form, the PSI logic defines five operational modes: Clear, Alignment, High Power, Post Experiment, and Stop. Definitions and permissions for each state are clearly identified. For example, the Alignment mode means that a low-power laser can be present in the hall. There is no prompt radiation hazard, but there are laser-associated hazards. Authorized personnel are allowed in the hall with appropriate personal protection equipment: dosimeters (activation and conventional sources may be present) and laser goggles.

Instead, in the High Power mode the experiment is running. Access to the experimental hall is not allowed. All doors to the hall are locked by the PSI. If a door is forced open, the PSI reacts by shutting the safety gate valves of the laser beam transport, thereby stopping the laser from propagating to the hall. In Post Experimental mode, entrance to the hall is permitted only to the Radiation Officer or a designated deputy. They sweep the hall, assess the residual radiation risk level, and set the system to Clear mode by logging into the PSI.

The hardware safety components of the PSI are safety gate valves regulating laser propagation, locks, limit switches for the access doors, flashing lights, audio broadcasting, and visualization panels informing about the operational status of the halls (Fig. 13). Finally, beam-off buttons are in the halls and control rooms for quick reactions to emergencies. Pressing a beam-off button cuts off the laser beam immediately, effectively stopping the experiment and bringing the hall to its Stop safe state.

Fig. 13. View of an experimental hall. Alarms, panels, and detectors are highlighted.

The PSI system is operated by using a touch screen panel and trapped keys located in the control room. Individual loggings with specifically defined roles for operators are required for access to the system. The PSI directly connects to other systems, including the radiation monitoring system described in Sec. X and the laser control system.

X. RADIATION MONITORING

X.A. Area Monitoring

In terms of surveillance of the current radiological situation, the PSI system relies on an online dose rate monitoring information system (MIS). The conceptual design, developed by the RP group, was implemented under its supervision by a consortium of external companies^60–62 during the years 2019 to 2022. The MIS combines passive and active detector technologies to monitor prompt radiation levels, activation, and contamination in the workplace and surrounding areas. The detection systems were chosen considering both their capabilities and the intended measurement objectives, ranging from complex information needed in the high-occupancy areas (control rooms) to simple indicators providing early warnings in low-occupancy areas (plant rooms). Namely, three goals were identified: measuring prompt gamma and neutron dose rates, providing an early warning system, and measuring continuous radiation from activated materials.

Prompt gamma and neutron dose rates are measured in the high-occupancy areas (control rooms) or regularly accessed areas (corridors). A combined novel monitor for pulsed radiation, the LB-6419 (Refs. 63, 64, and 65) is used. It comprises a helium-based proportional counter for neutron detection and a plastic scintillator for beta and gamma detection. Also, the characteristics of the radiation field in some locations has allowed for the installation of photon detectors only. A wide-range pressurized ionization chamber⁶⁶ was selected for this purpose.

Detectors for early warning have been implemented in the low-occupancy areas (plant rooms) and in the close vicinity of the interaction chambers for work during the Clear mode (see Sec. IX), when production of ionizing radiation is not expected but can still occur in the extremely rare case of the failure of several layers of protection. To detect prompt gamma and electron radiation, inorganic scintillators with aluminum-garnet crystals (Y₃Al₅G₁₂:Ce and Lu₃Al₅G₁₂:Ce) were chosen, since these crystal materials have been proven to have excellent scintillating properties even in harsh radiation environments.⁶⁷ Instead, prompt neutron radiation is detected by a proportional ³He counter inside a polyethylene sphere moderator (25-cm diameter). Possible activation in the experimental halls instead is monitored via Geiger-Müller counters⁶⁸ positioned in areas where high activation levels are expected

Information from all the monitoring detectors is collected, archived, and made available in real time in all the control rooms and remotely. Two signalization thresholds are set for emergency situations: warning and alarm. These are displayed and audibly announced directly in the halls and contemporarily signaled to the PSI, which takes the appropriate actions. The MIS is complemented by hand-held contamination monitors available in each experimental hall, a high-purity germanium spectrometer for detailed analysis in the central radiation laboratory, and a hand-foot monitor located at the exit from the controlled areas. Last, the online measurements are supported by the passive detection system based on optically stimulated luminescence dosimeters (see Sec. X.C).

X.B. Personal Dosimetry

A two-dosimeter rule is implemented at ELI Beamlines. Employees and users are requested to wear an electronic personal dosimeter (EPD) and a passive dosimeter. The EPDs (Ref. 69), which are sensitive to gamma and beta particles and provide a real-time measurement of personal deep [Hp(10)) and surface (Hp(0.07)] dose equivalent, were chosen for this purpose. Additionally, EPDs from the same manufacturer, but sensitive to gammas and neutrons, are also available. All EPDs have an active warning system (audible alarm) and internal memory for data retrieval. Readout and data analysis of the EPDs are performed inhouse, while regular calibrations are performed at the Czech Metrology Institute⁷⁰ (CMI).

The EPDs are indeed capable of detecting a pulsed field (<1 µs and a repetition rate <1 kHz) (Ref. 69). Nevertheless, their reliability in pulsed radiation fields as demanding as those at ELI Beamlines is limited. Therefore, anyone accessing the radiation-controlled area is required to wear a film badge on their chest (reference point). This also fulfills the Czech legal requirement (see Sec. IV) that prescribes the use of certified and legally recognized dosimetry systems, film dosimetry being one of these. Film badges are prepared and readout at the Czech National Service for Dosimetry.

X.C. Area Dosimetry

Optically stimulated luminescence (OSL) dosimeters made of beryllium oxide (BeO) are used^71,72 for area photon dosimetry. BeO was chosen for its favorable dosimetric characteristics: high sensitivity to photon ionizing radiation, linear dose response over six orders of magnitude (µGy to Gy), and low effective atomic number (Z_eff = 7.2), which makes it a near tissue-equivalent material.⁷³ Several thousand chips, each measuring 4.7 × 4.7 × 0.5 mm³, have been purchased and are placed throughout the entire laser building: halls, corridors, control rooms, and service areas. Evaluation is done inhouse, annealing is performed in a laboratory chamber furnace,⁷⁴ and readout is done using a Lexsyg smart reader.⁷⁵ Batch calibrations were performed at the Czech National Radiation Protection Institute (SÚRO, acronym in Czech), an accredited calibration laboratory^76,77 and at CMI (Ref. 70), a calibration laboratory. One benefit to using passive solid-state detectors is that, contrary to active detection systems, protection against EMPs is not necessary. Implementation of a passive neutron monitoring system is scheduled for the near future.

Finally, long-term environmental radiation monitoring using BeO OSL dosimeters around the facility premises is in place. ELI Beamlines is, in fact, located in an urban area, surrounded by public, private, and commercial buildings. Ten measuring locations have been selected in the parking lot and neighboring streets (see Fig. 14). All dosimeters are evaluated quarterly. Systematic measurements started in 2018 after the initial measurement performed by the certified laboratory SÚRO (Refs. 76, 77, and 78). Also, since the underground water table is quite high in the area, a dedicated study of the environmental impact of long-term potential activation was performed, while, standard protocols are also in place to monitor possible sources of air, water, and ground contamination.

Fig. 14. ELI Beamlines buildings and surrounding areas. The red dots mark the 10 environmental measuring points.

XI. CONCLUSIONS

The worldwide proliferation of high-power, high-intensity laser accelerators has opened a new chapter in the field of RP. The existing international standards and best practices used in conventional accelerator facilities cannot be directly applied to this new kind of facility, and new ones have not been firmly established yet. This paper reviews the challenges encountered and the adopted solutions while designing and implementing the RP system at the high-power, high-intensity laser facility ELI Beamlines. It comprises more than a decade of work and dedication from the RP group. This work may serve as a reference or an inspiration for new installations within this rapidly developing field.

The high-power, high-intensity laser facility laser, ELI Beamlines, poses considerable RP challenges. Wide penetrations in walls and ceilings are present to transport lasers and to house supporting technologies weakening the shielding elements. Therefore, ad hoc local shielding has been designed taking into account a combination of hazards: ionizing radiation, EMPs, and fire. The source term descriptions, needed for Monte Carlo simulations, are not as well known as in conventional accelerators and are subject to high uncertainty, requiring a conservative worst-case scenario approach. Laser-target interactions create strong EMPs, potentially damaging radiation detector electronics, thus a combination of active and passive detectors needs to be used. The experimental halls are clean rooms, significantly limiting available choices for shielding materials and the monitoring detectors that can be used; granite and coated concrete are effective options. The ultrashort laser pulses produce bursts of prompt radiation too short (up to ∼femtoseconds) to be reliably detected by conventional active dosimeters. The ELI Beamlines RP group is active in the testing and characterization of existing and newly developed detectors in ultrahigh pulse dose rate beam facilities. Finally, robust and redundant interlock and radiation monitoring systems are in place as fundamental layers of protection.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Supported by the project Advanced research using high intensity laser produced photons and particles (ADONIS) CZ.02.1.01/0.0/0.0/16_019/0000789 from European Regional Development Fund (ERDF).