Can Tritium Monitoring and Control Requirements for DEMO Be Met by Existing Technologies?

Abstract The tritium inventory of future fusion power plants needs to be monitored in the fuel cycle for several reasons: to comply with limits imposed by environment and safety regulators, to adhere to practices required by nuclear regulators, and for process control purposes. Fulfilling all these requirements leads to a comprehensive list of locations in the fuel cycle where tritium monitoring needs to take place, each characterized by different measurement conditions and required accuracies. Meanwhile, existing tritium detection technologies all come with specific applicabilities such as accuracy, material phase, and ability to detect tritium in a continuous manner. These do not necessarily correspond to the required measurement conditions. As an example, one tritium detection technology will be matched up with the previously defined measurement conditions, which allows for the identification of gaps in the existing detection capabilities of this technology. This work leads to several recommendations, i.e., developments to expand the applicability of tritium detection technologies, experimental proposals to test detection techniques at more extreme conditions, and expansion of the regulatory framework regarding tritium handling and breeding. These developments are critical for a functioning tritium management and control system, and this paper outlines the first step in that process.


I. INTRODUCTION
Fusion power plants such as the DEMOnstration fusion power plant (DEMO) will use the fusion reaction between the hydrogen isotopes deuterium (D) and tritium (T) for energy production.Using tritium involves the same difficulties as working with the other hydrogen isotopes, hydrogen and deuterium, such as metal embrittlement.Additional handling challenges are also introduced due to its radioactivity and potential nonproliferation issues caused by tritium's capability to increase the sophistication of nuclear weapons programs.
Because of these challenges, the DEMO fuel cycle will need a tritium management and control system.This system will account for tritium inventories and flows at all locations in the fuel cycle and at all times to ensure that the following four requirements are met: 1. Safety: Reduce the amount of tritium that is potentially released in accident scenarios to specified limits, maintain concentrations and amounts of substances at safe values to avoid dangerous situations such as creating explosive mixtures, and monitor and minimize worker radiation exposure.

Environmental protection:
Limit the amount of tritium (gas, liquid, solid) released during both normal operations and accidents.
3. Nonproliferation: Reduce the likelihood of material diversion/theft and identify its occurrence by specifically monitoring for any diversion of tritiated material that might be used for purposes related to nuclear weapons.4. Plasma and process control: Ensure stable compositions of plasma and gas mixtures in fusion reactors and monitor processes to detect leaks promptly and enable adequate control of processing systems within the fuel cycle.
These requirements each impose the need for tritium detection at different locations in the fuel cycle and at different time intervals. 1Current experience with a tritium management and control system exists only from experimental nuclear fusion reactors like Joint European Torus (JET).However, for various reasons, simply adopting this system for a DEMO fuel cycle is not practical.JET operates in pulse mode and uses small quantities of tritium, allowing for most measurements to take place in batch mode.Moreover, most of the accountancy operations that accompany any operation in the JET active gas-handling system are done by hand.In contrast, DEMO will operate in (near) continuous mode, requiring suitable detection methods, while the implementation of automated systems is necessary to reduce the staffing requirements and likelihood of human errors.Further, a DEMO plant will also both consume and produce significant amounts of tritium, and some of that which is produced will be exported to other fusion power plants.The large amounts of tritium present will provide a larger risk of diversion of relevant quantities for proliferation activities, and the additional complexity of future fusion power plants may make material diversion easier to conceal.Although there is no regulation from the International Atomic Energy Agency (IAEA) regarding tritium accountancy right now and control is imposed on a national basis, it can be expected that this may change.It should however be noted that the proliferation threat posed by tritium is that of improvement of existing nuclear weapons programs rather than the spread of nuclear weapons to new countries.Without weaponsgrade nuclear material, the possession of tritium is not a proliferation concern, although it may reduce the breakout time of nation states.
The outline of this paper is as follows.First, the different requirements will be discussed in more detail, explaining the purposes they serve and the locations in the fuel cycle where these measurements need to take place.The locations all come with varying sets of conditions that need to be categorized, and the applicability of available detection technologies to these condition categories will be discussed.This paper concludes with recommendations for improvements in the applicability of detection technologies, how these can be tested, and for which conditions new technologies will need to be developed.

II. TRITIUM DETECTION LOCATIONS
There are four different purposes for which tritium accounting and monitoring are needed in the DEMO fuel cycle, each with its own requirements regarding accuracy, frequency, and locations where tritium detection needs to take place (Fig. 1).

II.A. Safety
As there is no location yet for DEMO, local safety agencies cannot be consulted regarding their requirements and regulatory processes.However, the two key goals will be first, the prevention and mitigation of accidents and second, the monitoring of radioactive material releases and doses to workers and the public.
The former can be achieved through appropriate plant design, including application of safety case methodologies, and limiting the tritium inventories in subsystems.These measures will reduce the source terms of tritium in an accident scenario.Radioactive releases and worker doses can be achieved using in-plant radiation monitoring such as ionization chambers for tritium and gamma and alpha detectors for other radiation releases, personal dosimeters, and bioassay sampling of workers.Other measures, such as limiting the airborne tritium inventories in each room and a zoning system for ventilation, aid the achievement of both goals. 3

II.B. Environment
Similar to safety agencies, it is not known yet which local environment agency needs to be consulted regarding tritium emissions from a DEMO plant.These emissions will constitute gaseous discharges of T 2 , HT, DT, and HTO vapor; liquid discharges of HTO and organically bound tritium compounds; and solid waste containing tritium.Other discharges that need to be monitored for radiological reasons are discharges of activated plasmaenhancement gases and activated products from sputtering and corrosion, especially tungsten dust from plasmafacing components.
Gaseous discharges can be detected using real-time ionization chambers located in the stack (stack monitors),

564
TIJSSEN and BUTLER although the accuracy of these instruments is typically not considered sufficient for quantitative monitoring.Instead, ionization chambers provide an indicative measurement and are linked to recording and plant interlock and alarm systems.For quantitative monitoring, emissions can be further monitored using tritium-absorbing gel packs that are routinely replaced and measured using scintillation methods to give accurate emission data.Vegetation patches can also be used to monitor local absorption of tritium by crops.
Sampling and scintillation analysis of all liquid outlets of the plant can be undertaken, with samples being taken to an on-site or off-site laboratory for scintillation analysis.Solid waste can similarly be sampled and surveyed before being taken away for controlled disposal in appropriate facilities.

II.C. Nonproliferation
Tritium itself is not a fissile material, but it can be used to enhance the performance of nuclear weapons.As such, Ontario Power Generation has made an agreement with the United Kingdom that all tritium that is supplied to JET is being used for peaceful nonexplosive uses of nuclear energy only. 4As such, inventory audits need to take place annually, even though tritium does not fall under the Nonproliferation Treaty as defined by the IAEA.
For now, since there is no specific guidance regarding tritium accountancy in fusion reactors, only assumptions on future regulations can be made, drawing from experience from the fission industry and other tritiumhandling facilities.It can be assumed that for nonproliferation purposes, all tritium quantities both entering and leaving the plant will need to be monitored.These include all tritium quantities that are imported from elsewhere; produced and extracted from the breeder blankets; burned in the fusion reaction; and exported to elsewhere or discharged as waste in either gaseous, liquid, or solid form.These quantities need to be measured at least annually and will be reported to the regulators.It can be expected that most of these measurements, such as the amount of tritium imported or exported, will take place in Fig. 1.Diagram of subsystems in the fuel cycle, 2 with colors indicating for which purpose tritium measurements need to take place.Blue: nonproliferation; green: environment; red: safety; orange: plasma composition control.General process control measurements can be expected to take place in almost every location and are hence not displayed here.batch mode.Moreover, a (planned) shutdown of the plant for maintenance will provide an opportunity to pull back tritium from some subsystems and perform accountancy on these.However, other processes such as the burnup and production rate of tritium in the fusion reaction can be determined only indirectly, for example, using neutron diagnostics. 5,6Another challenge will be to determine the "losses" of tritium by being trapped in areas such as the first wall and breeder blanket, although these are expected to remain relatively constant once equilibrium is reached after the start-up of the plant.Finally, the radioactive decay of tritium needs to be considered for the total inventory of the plant.
To make sense of these measurements, analogous to practice with fissile materials in fission power plants, it needs to be defined what constitutes a significant amount of tritium to be diverted and during what time frame. 3hese accountancy activities need to occur with enough precision and high enough frequency to detect any tritium losses that exceed these limits.A useful and widely used tool for this purpose is the introduction of mass balance areas, i.e., areas where the quantity of tritium transported into or out of is determined and of which the total tritium inventory can be measured when required.

II.D. Process Monitoring and Control
Process control is required to operate the plant safely and efficiently and to ensure that gas mixtures are of the right compositions.It can be assumed that process control data comprise the information provided to the control room to serve as indicators as to whether processes are occurring as normal and all equipment is functioning correctly or whether there is a fault somewhere in the plant.Further, the plasma composition needs to be controlled to ensure that DT fusion occurs at its optimum for power output.This will depend on the correct functioning of isotope separation systems and fueling systems, such as pellet and gas injection.

III. MEASUREMENT CONDITIONS
Considering all these requirements leads to a large collection of measurement locations in the DEMO fuel cycle, which are each characterized by a set of measurement conditions.To find a suitable tritium detection technology for every location, these measurement conditions need to be categorized first.This is a challenging task as there are many different flow rates, temperatures, matter phases, tritium concentrations, and external factors such as magnetic fields and radiation in the DEMO fuel cycle.Further, some of these flow rates are still unknown.Another important factor is whether the measurement is on a subsystem that works in batch mode or is on a continuous flow.In the case of continuous flows between subsystems, the tritium detection must be continuous, either in-line (all flow is analyzed) or on-line (part of the flow is analyzed and then returned to the system), giving immediate results.Although most measurements will fall into this category, there will be some subsystems where static or off-line measurements (a sample is taken away to be analyzed, and measurement is not instantaneous, for example, calorimetry) can take place, such as storage systems.
To overcome these difficulties, all system blocks in the fuel cycle have been examined, considering the inflow and outflow of all gases and liquids while also eliminating duplicate mass flows.This results in 46 different connections between system blocks.The conditions occurring at each connection between the system blocks have been categorized using Table I.
The flow rate has been categorized logarithmically, as a large range of magnitudes is considered and measurement technologies are usually suitable only at flow rates of a specific order of magnitude.The temperature has been categorized into "ambient," "warm," and "cold" as the majority of processes take place at ambient temperatures and experiments to verify the detection technologies will not need special controls in the ambient range, while they will require heating or cooling in the "warm" and "cold" ranges, respectively.The divide between "warm" and "hot" has been chosen to coincide with the boiling point of water, as the presence of water vapor involves additional complications for tritium measurement.Similarly, the "cryo" temperature has been chosen to coincide with the freezing point of hydrogen.The tritium concentration has been split into "high," "medium," and "low" as equipment that is suitable to detect very low concentrations will become oversaturated at high concentrations, and vice versa, equipment that can detect high concentrations does not have the accuracy to determine very low concentrations.Further, there will be gas streams in the fuel cycle that are required to be (close to) pure tritium, some gas streams with a DT mixture close to 50:50, and others where tritium should not be present at all.The other categories are based on discrete properties such as the phase of the matter investigated and the presence of ionizing radiation or magnetic fields.
The conditions occurring at some connections may fall into the same category while there are also combinations of conditions that do not occur at all.As a result, there are 17 different sets of conditions identified.For each of these conditions, the most suitable tritium detection technology needs to be found.

566
TIJSSEN and BUTLER • TRITIUM MONITORING AND CONTROL REQUIREMENTS FOR DEMO

IV. DETECTION TECHNOLOGIES
Previously, we studied the suitability of tritium detection techniques at DEMO fuel cycle relevant measurement conditions, weighing them against current technological readiness and the possibility of further development, applicability to use with tritium and under the conditions expected, accuracy, precision, cost, and commercial availability.This work has produced a short list of seven detection technologies that will be discussed briefly below.

IV.A. BIXS
Beta-Induced X-ray Spectroscopy (BIXS) depends only upon the beta particles produced from a source.Hence, the method can be used to analyze tritium in a wide range of chemical states.BIXS can be used for water samples, although with a lower sensitivity than for gas, and on solids.One advantage of using BIXS is that it can be used under dynamic conditions in liquids, 7 as opposed to, e.g., liquid scintillators.However, BIXS is not commercially available and suffers from memory effects. 8

IV.B. Scintillation Counter
Liquid scintillation counting, which uses a scintillation cocktail that is mixed with the tritiated liquid, is commonly used because of its low limit of detection but can be used only in batch mode.Solid scintillators, such as fibers, can be used for real-time monitoring of tritiated liquids. 9Gaseous scintillation detection systems are not commercially available but have been tested for on-line static measurements. 10owever, this was more than 25 years ago and has not been followed up since.

IV.C. Raman Spectroscopy
Raman spectroscopy is a method based on the inelastic Raman scattering of laser light off a sample, which can be in gaseous, liquid, or solid form.Raman spectroscopy can distinguish among all six hydrogen isotopologues due to the relatively large mass differences, making it suitable for qualitative measurements, and can be calibrated for taking quantitative measurements.However, it is unsuitable for monatomic species such as helium.Measurement times are typically of the order of seconds, making it suitable for continuous measurements. 11However, to increase accuracy, usually the average of several spectra is taken, increasing the measurement time.Raman spectroscopy is commercially available although developments are needed for gas phase detectors and to ensure tritium compatibility.

IV.D. Mass Spectrometry
Mass spectrometers are generally used for quasistatic gaseous measurements, where a small amount of gas is sampled by breaking the particles up into ions and analyzing their mass/charge ratio.Mass spectrometry has the potential to measure gas compositions up to pure tritium with no memory effect, and there is high confidence that this technique can be used to measure continuously in fusion-relevant conditions with good reliability.Recent developments have made it possible to sample gas streams up to 30 bars and 200°C.Mass spectrometry measurements can be combined with other detectors such as flowmeters to provide information in terms of g/m 3 or mol/m 3 .Further customization is needed toward the calibration of fraction patterns containing tritium of, for example, CQ 4 , and to address the issue of line overlap, which is a primary concern for hydrogen isotopologues as their fragments occupy a very narrow mass range with very small mass differences between some species.

IV.E. Gas Chromatography
Gas chromatography is a separation technology and needs to be combined with another technology to act as detector.For example, using knowledge of characteristic retention times and monitoring physical properties of the elution allow for identification of the eluent.
Micro-gas chromatographs work faster, with a retention time of several minutes, which is not quick enough yet for continuous measurements. 12Moreover, because of the use of a carrier gas, this method cannot be used in a continuous manner as this gas cannot reenter the gas stream.

IV.F. Calorimetry
Calorimetry is a tritium detection method that is commonly used for solid storage systems, as measurements can typically take hours.By measuring the temperature increase in the sample due to radioactive decay, the total amount of tritium can be determined.As a result, it needs to be combined with pressure-volumetemperature calculations to determine the concentration of tritium in a mixture.Calorimetry measurements generally need to be performed in tandem with a reference volume.

IV.G. Ionization Chambers
Ionization chambers produce a small electric current with its magnitude a measure of the amount of radiation present.Ionization chambers are very sensitive to memory effects, moisture, and gas pressure.Their use in tritium laboratories is common and widespread, for example, for gas discharges and air monitoring; however, because of these limitations, they are usually used only as indicators and not for quantitative measurements.

IV.H. Other Technologies
Other possible candidate technologies for solid measurements such as on first wall materials are laserinduced breakdown spectroscopy (LIBS) and laserinduced ablation spectroscopy (LIAS) (Refs.13 and  14).Here, a high-energetic laser produces a plasma of particles released from the solid material, which can be

V. GAP ANALYSIS AND RECOMMENDATIONS
The properties of the measurement technologies discussed are summarized in Table II, focusing on whether the technology can be applied in a continuous manner and which state of matter can be detected.
Comparing the measurement conditions at which tritium detection is required with the available technologies shows that in general, the main difficulty lies in finding technologies that are suitable for in-line measurements, with real-time results.Raman spectroscopy is a very promising candidate for these conditions, having the additional advantage that it is suitable for all three phases of matter.BIXS, although having the same advantages, is probably of a technological readiness level that is too low for application in DEMO right now.However, DEMO is still many years away, and it may be worth developing BIXS further as it could offer many applicabilities.Further, with developments toward higher resolutions, mass spectrometry could be exploited for composition analysis.Other technologies, such as calorimetry, gas chromatography, and ionization chambers, are of a more mature readiness level and can be applied in those conditions for which they are already being used.However, modifications might be desired to widen their applicability, especially to continuous modes or higher flow rates.
Future experimental activities will be necessary to test the applicability of these technologies to tritium detection at various conditions.Currently, at United Kingdom Atomic Energy Authority, an experiment is being designed with this aim.As gaseous conditions make up most conditions present in the DEMO fuel cycle, these conditions are prioritized in this rig.However, it is important to keep in mind that the detection of tritium in especially solids and detection in the presence of ionizing radiation or magnetic fields are at a lower technology readiness level and will need to develop further before being applicable to the DEMO fuel cycle.

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

Funding
This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (grant agreement number 101052200 -EUROfusion).Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission.Neither the European Union nor the European Commission can be held responsible for them.

ORCID
Teuntje Tijssen http://orcid.org/0000-0002-7899-8697 These technologies can be used for determining the thickness and composition of deposition layers on the first wall.
SCIENCE AND TECHNOLOGY• VOLUME 80 • APRIL-MAY 2024 analyzed using spectroscopy.