The Joint European Torus (JET) reactor located at Culham Centre for Fusion Energy (CCFE) is currently the world’s largest magnetically confined plasma experiment for the realisation of nuclear fusion to generate low-carbon electricity (CCFE, 2020). Fusion scientists have been striving to exceed plasma energy breakeven, and thus show that nuclear fusion is a viable means for humanity to generate almost unlimited power since the 1940s (Seife, 2008). In an era of ever-increasing environmental awareness (Delis and Iosifidi, 2019), the prospect of a reliable and environmentally sustainable energy source such as nuclear fusion has gained international funding and interest, culminating in the most ambitious energy project in history – The International Thermonuclear Experimental Reactor (ITER). The ITER project intends to be the first fusion device to produce net energy and thus pave the way for the future fusion power plants (ITER, 2020) which are known as DEMO-class power stations (Coleman et al., 2019). JET remains at the forefront of research into nuclear fusion despite being operational for over 30 years and is now providing a platform for testing systems, materials and designs for ITER (CCFE, 2020).
Fusion reactions for terrestrial energy production involve the reaction of nuclei by the strong nuclear force on light elements; specifically, the isotopes of hydrogen (Morse, 2018). The difference between the mass of the reactant and product nuclei results in enormous energy release due to mass-energy equivalence according to Einstein’s famous equation, E =mc^2. D-T fusion, the reaction between deuterium (D) and tritium (T), gives rise to a 17.6 MeV energy release, and so is the most efficient fusion reaction (Morse, 2018). Therefore, the achievement of a sustained D-T fusion reaction in a reactor is the goal of current fusion research (Tanabe, 2017b). Deuterium fuel is abundant, with an estimated inventory of 1.38 x 1043 atoms in seawater which can be extracted by electrolysis. If this inventory were to be exhausted by nuclear fusion, it is estimated that the deuterium supply would last for 25 billion years (Morse, 2018). It is expected that energy demand may triple by 2050 (Siirola, 2014), so the prospect of an almost unlimited energy source such as nuclear fusion provides an attractive alternative to finite fossil fuels such as crude oil and natural gas, which account for 60% of the world’s current primary energy demand (Norouzi et al., 2020).
Tritium is radioactive and decays with a half-life of 12.323 years or 5.47 % per year. It decays by beta decay to 3He producing an electron and an antineutrino. Tritium is produced when cosmic rays interact with nitrogen in the atmosphere. Therefore, tritium exists in a production/destruction equilibrium at an approximately constant concentration of 10-18 in natural hydrogen, known as the tritium unit (TU) (Tanabe, 2017a). This means that the natural steady-state global inventory of tritium is about 7.3 kg (Ojovan et al., 2019). It is estimated that ITER will require 1.1 kg of tritium annually from 2035 for its expected 12 year operation period (Pearson et al., 2018). Additional DEMO-class fusion power plants designed post-ITER may require up to 50 kg of tritium at start-up (Pearson et al., 2018).
It is evident that tritium must be produced artificially for D-T fusion to be viable, because tritium does not occur readily on Earth in large quantities (Tanabe, 2017a). Approximately 130g of tritium is produced annually in a typical CANDU (CANada Deuterium Uranium) fission reactor (Pearson et al., 2018) from neutron activation of 6Li, known as tritium ‘breeding’ (Giegerich et al., 2019). The tritium is removed from the tritiated water using a Tritium Removal Facility (TRF). This complex process is operational in two locations globally; Canada and South Korea, although there are plans for a further TRF in Romania (Pearson et al., 2017). Tritium ‘breeding’ also occurs in the fusion reaction when neutrons escaping the plasma interact with 6Li contained within the blanket wall of a tomahawk reactor (ITER, 2020). The tritium produced from these methods, in combination with approximately 20 kg of global tritium inventory (ITER, 2020), will be available for the operation of ITER (Pearson et al., 2018). To ensure that the global tritium inventory is not depleted, ITER and future DEMO-class fusion reactors must demonstrate tritium self-sufficiency by ‘breeding’ tritium in enough quantities to compensate for the 5% decay rate and the required start-up inventories (Coleman et al., 2019). This will ensure that there is enough tritium fuel for nuclear fusion to become a reliable component of the global energy mix.
The world record power output from a fusion device was achieved by JET in 1997 is 16 MW using a plasma volume of 100 m3 with a fusion energy gain factor (Q) of Q=0.67 (Keilhacker et al., 2001). ITER intends to be the first nuclear fusion reactor to produce net energy, using a plasma volume of 840 m3, and is expecting Q≥10 (ITER, 2020). Lessons learned from ITER will be used to design future DEMO-class power plants, where it is expected that 2 GW of thermal power will be produced, meaning that the order of 102 kg of hydrogen isotopes are expected to be recycled each day (Shaw and Butler, 2019). It is therefore clear that the design, operability and scale-up potential of hydrogen isotope separation processes to recover tritium are an important aspect of the delivery of economical nuclear fusion.
JET Active Gas Handling System
Tritium (T2) and deuterium (D2) are recycled at JET using an Active Gas Handling System (AGHS). The purposes of the AGHS are the following: Continuously evacuate equipment connected to the torus and the torus of active gases (Lässer et al., 1999). This prevents the accumulation of substances, which may create safety issues. This purpose is fulfilled by the Cryogenic Forevacuum (CF) and Mechanical Forevacuum (MF).
Separation of the pumped active gases (Lässer et al., 1999). In the torus there are protium (H), deuterium (D) and tritium (T) atoms, so the six main hydrogen molecules present in the active gas are; H2, D2, T2, HD, HT and DT (Yamanishi, 2017). Hydrocarbon impurities are formed by the interaction of the plasma with atomic tritium with carbon on the torus reactor wall. Helium is also present because it is added for power control density (Lässer et al., 1999). Other species such as water, oxygen, argon and nitrogen may also appear in trace quantities due to small leaks (Lässer et al., 1999). The hydrogen isotope separation processes utilised at JET are Cryogenic Distillation (CD) and Gas Chromatography (GC), which will be discussed in detail later in this section.
Detritiation of impurities and recovery of tritium (Lässer et al., 1999). As discussed in section 1.1, tritium is scarce and should be recovered to avoid depleting global inventories. This purpose is fulfilled by Impurity Processing (IP) after the torus has been evacuated, as well as the hydrogen isotope separation processes employed, CD and GC. Isotopic preparation of hydrogen mixtures into T2 and D2 (Lässer et al., 1999). This is performed by the hydrogen isotope separation processes, CD and GC.
Storage of T2 and D2 (Lässer et al., 1999). The JET AGHS uses Intermediate Storage (IS) to store detritiated species before they are injected into separation processes, and Product Storage (PS) to store T2 and D2 products. The method of storage of the gas is in uranium beds (Lässer et al., 1999). Resupply of T2 and D2 back into the torus through the Gas Introduction box (GI) (Lässer et al., 1999).
Discharge of detritiated compounds safely into the atmosphere through the Exhaust Detritiation (ED) system (Lässer et al., 1999). Regulatory approval had to be obtained following with the UK Radioactive Substances Act of 1993 for JET to hold, accumulate and discharge tritiated waste (Patel et al., 1999), with the annual limits of 90 TBq for tritium oxides, and 110 TBq for other all tritium compounds (Bell et al., 1999).
To fulfil all purposes safely with minimal operator and environmental risk (Lässer et al., 1999). The safety case for JET (Bell et al., 1999) states that during D-T operation, the annual dose to classified worked should be less than 5mSv, and to all other staff, less than 1mSv. All necessary precautions are taken to protect against radiation, and secondary containment is fitted on all systems (not including ED). Some process units are configured with tertiary containment and a ventilation system is also in the active gas handling building (Lässer et al., 1999). This ensures that operators and staff are kept safe from radiation by ensuring tritium and other radioactive substances are contained under a contingency scenario.
It is important to critically review the hydrogen isotope separation processes at JET because it is expected that the methods used in the JET AGHS will be unsuitable for larger fusion reactors such the ones proposed for ITER and DEMO-class power plants. This is the case because the methods used in the JET AGHS are expensive, complicated, energy intensive and require a lot of attention from operators (Smith et al., 2015). ITER and the future DEMO-class fusion power plants will include integrated hydrogen isotope separation systems with the main focus of recovering tritium due to its scarcity (Tanabe, 2017a) and to fulfil the requirement for tritium self-sufficiency (Coleman et al., 2019), as discussed in section 2.1. Therefore, this research aims to review and assess the hydrogen isotope separation processes used at JET (specifically CD and GC) and provide an insight into whether these systems or alternative options should be carried forward into ITER and DEMO-class power plants.
Cryogenic distillation (CD) columns are similar to conventional distillation columns but are operable at very low temperatures. A separation is performed by capitalising upon the differences between vapour pressures of the individual species in the hydrogenic mixture. Use of an external multi-layer refrigeration cycle to provide the cryogenic temperatures is preferred, meaning the efficiency of CD processes is hugely dependant on cold recovery systems in place. Increasing the pressure in a CD column would mean that the separation could be performed at closer to ambient temperature. However, the maintenance of high pressures is costly, requires more capital expenditure and as relative volatility generally decreases with pressure, using higher pressures may render the separation more difficult (Sinnott and Towler, 2019b). Operating a distillation column at cryogenic temperatures is energy intensive and requires large capital investment. Furthermore, cryogenic temperatures make operational procedures, start-up and shutdown more difficult. Despite these challenges, CD is regarded as the separation process of choice for large-scale hydrogen isotope separation due to their high separating performance, compact unit size, negligible tritium permeation, and ability to operate over a range of feed flow rates and compositions (Yamanishi and Kinoshita, 1984).
CD technologies were first designed at the Los Alamos National Laboratory in 1979, where the design of 4 interlinked fractional distillation columns operating at cryogenic temperatures was produced to handle a mixed hydrogenic stream. The CD system was designed to produce high purity feed streams for thermonuclear reactors (Bartlit et al., 1979). Early studies involving mathematical model development and computer simulations allowed experiments to be performed by Yamanishi and Kinoshita in 1984, which gave indispensable design information for CD columns on packings, HETP values, pressure drop and operation (Enoeda et al., 1989). Further experiments at the Tritium Process Laboratory in the Japan Atomic Energy Research Institute (TRL/JAERI) have also shown the reliability of CD technologies (Iwai, Yamanishi, et al., 2002). Safety studies assessing the impacts of the loss of cryogenic cooling (Iwai, Nakamura, et al., 2002) have been conducted to ensure CD columns can be operated safely and to ensure tritium can be recovered under contingency scenarios in future applications at ITER and DEMO. The ability to process large volumes of gas while maintaining high purity product composition is a significant advantage of CD when hydrogen isotope separation processes will need to be scaled up for future fusion applications (Draghia et al., 2016).
In the JET AGHS, a hydrogenic stream is produced after impurity processing, and consists of H2, HD, HT, D2, DT and T2 (in rank order of volatility with H2 being most volatile). This hydrogenic stream is passed to three CD columns housed inside a Process Cold Box (PCB) which is actively cooled by a 200 W helium refrigeration cycle. The distillation is performed at temperatures in the range of 20-30 K to produce a stream of pure H2 which is discharged to a stack, as well as D2 and T2 streams for recycle (Bainbridge et al., 1999). Typically, a multicomponent mixture of n species would require n-1 columns to separate into pure products (Sinnott and Towler, 2019c), but only three columns are used in the JET AGHS due to strategically placed equilibrators which remove HT and promote dissociation of DT (Bainbridge et al., 1999). During the JET DTE1 experimental campaign, the CD system delivered H2 and D2 with average tritium concentration ≪ 1 ppm. This meant that H2 could be safely discharged through the stack. Transfer of a pre-enriched tritium stream to Gas Chromatography (GC) separation processes was also possible due to tritium enrichment up to 35 % in the right conditions (Bainbridge et al., 1999).
Operational experience from the CD system at JET will provide a valuable insight when hydrogen isotope technologies are selected for ITER and DEMO (Lawless et al., 2017). CD offers an attractive method to fulfil the requirements for hydrogen isotope separations at ITER and DEMO, despite large capital and operational costs. This is because it is able to deliver the high purities required whilst being able to accommodate for variations in feed compositions and flow rates even as the systems size is scaled up (de Haan and Bosch, 2013). However, the supply of helium for the cryogenic refrigeration systems is an issue because it is a globally scarce element. Cryogenic applications made up approximately 30% of US helium consumption in 2018 (Malinowski et al., 2018), so for ITER a centralised helium cryogenic plant is being constructed to ensure that the necessary cryogenic temperatures of 20-30 K can always be provided (AirLiquide, 2020).
The principle of chromatographic separation relies upon the different species in a fluid stream having different adsorption equilibria relative to a solid phase, and thus different retention times (Sinnott and Towler, 2019a). Gas Chromatography (GC) for hydrogen isotope separations involves passing a fluid stream over a bed of absorbent material, which is typically a palladium alloy. Palladium is used because it has large isotopic effects on hydrogen absorption, making it ideal for effecting a separation (Watanabe et al., 1998).
There are a variety of different chromatographic techniques available; frontal, elution, displacement and self-displacement. It has been found that displacement chromatography is more useful for the separation of a small amount of deuterium or tritium from hydrogenic gas mixtures (Fukada and Fujiwara, 2000). The AGHS at JET employs a displacement GC technique, by passing hydrogenic gas through a GC column packed with 18 to 20% palladium deposited on porous α-Al2O3, which can absorb 1.57 Pa m3 per gram of packing (Botter et al., 1990). The GC column is filled with Helium and heated to 320 K to begin operation, and then hydrogenic gas is fed into the column. Lighter hydrogen species absorb into the bed, while heavier hydrogen species prefer to stay in the gas phase. This means that as hydrogenic gas is forced through the column, a tritium rich region expands, absorbing tritium from the hydrogenic gas mixture. Protium displaces deuterium and tritium as it is the lightest component of the hydrogenic gas mixture, hence the name of the chromatographic separation. The products from the GC column therefore emerge in the following sequence; He, T2, D2 and finally H2. The palladium bed is then regenerated by heating the column to 473 K and then circulating helium to desorb protium. During the DTE1 experimental campaign, tritium quality of 99.9% was achieved under batch operation (Lasser et al., 1999).
Effective separations of high purity can be delivered using GC, with low operating cost, especially in comparison with CD. The method is also fairly straightforward to operate and maintain reliably (Lasser et al., 1999). However, the process is batch, which may be a significant issue for ITER and DEMO-class power plants which will have increased separation requirements. Continuous processes are being investigated, but require more complex plant and equipment involving moving beds (Owens, 2015). Furthermore, GC systems require palladium for their packings, and there are concerns regarding cost, availability and degradation – an issue which will be compounded as hydrogen isotope separation processes must be scaled up for ITER and DEMO.