Hydrogen Inventory Simulations for PFCs (HISP)

This paper introduces HISP, an open-source simulation tool used to demonstrate that baking is the most effective method for removing tritium from ITER's plasma-facing components, significantly outperforming Glow Discharge Conditioning and low-power deuterium pulses which had minimal impact on the final inventory.

The Gist

The Big Picture: The "Tritium Ticking Time Bomb"

Imagine a fusion reactor (like the giant ITER machine) as a high-performance car engine. To make it run, it burns a special fuel called Tritium (a radioactive form of hydrogen).

The problem is that this fuel is sticky. When the engine runs, tiny particles of the fuel get stuck in the walls of the engine (the "Plasma Facing Components"). If too much fuel gets stuck, two bad things happen:

1. Safety: It becomes radioactive and dangerous.
2. Efficiency: You run out of fuel because it's stuck in the walls instead of burning in the engine.

The International Thermonuclear Experimental Reactor (ITER) has a strict rule: You can only have 1 kilogram of this sticky fuel stuck in the engine at any one time. If you go over, you have to stop and clean it.

The Problem: How Do We Clean It?

The scientists in this paper asked: "What is the best way to clean the sticky fuel off the engine walls?"

They know there are a few ways to do it, like baking the engine, blowing gas through it, or running a low-power test drive. But they can't just guess; they need to know exactly how much fuel will be removed by each method to stay safe and efficient.

The Solution: The "HISP" Simulator

To solve this, the team built a digital tool called HISP (Hydrogen Inventory Simulations for PFCs).

Think of HISP as a super-advanced video game simulator for the reactor.

• The Input: It takes data from other complex physics programs (which tell it how the fuel hits the walls).
• The Process: It breaks the reactor walls into tiny, manageable "slices" (like slicing a loaf of bread). It simulates how the fuel sticks to each slice over time.
• The Output: It tells the engineers exactly how much fuel is stuck in the walls after a week of running.

The Experiment: Three Cleaning Scenarios

The researchers ran three different "what-if" scenarios in their simulator to see which cleaning method worked best. They imagined a 2-week cycle of the reactor running.

1. Scenario A: "Do Nothing"
   • The reactor runs for 10 days, then they take a short break, then run again. No special cleaning is done until the very end of the 2 weeks.
2. Scenario B: "Just Glow"
   • Same as above, but during the short breaks, they use a technique called Glow Discharge Conditioning (GDC). Imagine this like using a gentle, cold plasma "wind" to blow some of the sticky fuel off the walls.
3. Scenario C: "Capability Test"
   • Same as above, but they also run a few "low-power test drives" (using Deuterium instead of Tritium) during the breaks. This is like running the car engine at idle to shake some of the dirt loose.

The Surprising Results

After running the simulations, here is what they found:

1. The "Sticky Tape" Effect (Boron vs. Tungsten)
The reactor walls are made of a hard metal called Tungsten (like a steel pan), but they are often coated with a thin layer of Boron (like a non-stick coating).

• The Finding: Even though the Boron layer is incredibly thin, it acts like super-sticky tape. It trapped about 80% of all the fuel! The Tungsten metal only held a tiny fraction.
• Analogy: Imagine trying to clean a floor. The wood floor (Tungsten) is clean, but there's a thin layer of honey (Boron) on top. The honey catches 80% of the dirt. You don't need to scrub the wood; you need to scrub the honey.

2. The "Oven" Wins (Baking)
The most effective cleaning method was Baking.

• What it is: Heating the entire reactor walls up to 220°C (428°F) for a week.
• The Result: This was like putting the engine in a giant oven. It baked the sticky fuel off the walls.
  • It removed 88% of the fuel stuck in the Tungsten.
  • It removed 30% of the fuel stuck in the Boron.
• Why not more? The Boron is so sticky that even the oven heat couldn't get all of it off. Some fuel is trapped so deep inside the Boron that it needs even hotter temperatures to escape.

3. The "Wind" and "Idle" Didn't Help Much
The other methods (Glow Discharge and Low Power pulses) were okay, but not great.

• They removed a little bit of fuel (about 10–23%), but because the Baking method was so much more powerful, adding these extra steps didn't change the final result very much.
• Analogy: If you have a muddy car, you can try to blow the mud off with a leaf blower (Glow Discharge) or drive it through a puddle to shake it loose (Low Power pulses). But if you just take it through a high-pressure car wash (Baking), that's what actually gets it clean. The leaf blower barely made a difference.

The Bottom Line

The scientists concluded that for the ITER reactor:

1. Baking is King: Heating the reactor is by far the best way to remove the dangerous radioactive fuel.
2. The "Sticky Tape" is the Enemy: The thin Boron layers are where the fuel hides. We need to focus our cleaning efforts there.
3. The Simulator Works: Their new tool, HISP, successfully predicted these results. It's a "proof of concept" that helps engineers plan how to keep the reactor safe and running efficiently without getting stuck with too much fuel.

In short: To keep the fusion reactor safe, don't just blow on it or idle it. Turn up the heat and bake the walls!

Technical Summary

1. Problem Statement

Managing tritium (T) inventory in magnetic confinement fusion reactors is critical for both fuel recovery and radiological safety. The International Thermonuclear Experimental Reactor (ITER) has a strict in-vessel tritium inventory limit of 1 kg, with a specific operational limit of 700 g for Plasma Facing Components (PFCs) after accounting for cryopumps and measurement uncertainty.

To maintain safety and efficiency, ITER must employ various tritium removal strategies, including:

• Baking: Heating the reactor to ~220°C (scheduled for Long Term Maintenance).
• Glow Discharge Conditioning (GDC): Using deuterium plasma during Short Term Maintenance (STM).
• Low Power Deuterium (DD) Pulses: Using pure deuterium plasmas during operation.

While these techniques have been demonstrated in smaller devices (JET, WEST), their extrapolation to ITER is uncertain due to geometric differences. There is a lack of computational tools capable of bridging the gap between plasma edge codes (which provide flux data) and hydrogen transport codes (which model inventory evolution) to predict the effectiveness of these removal strategies on a reactor scale.

2. Methodology: The HISP Framework

The authors developed HISP (Hydrogen Inventory Simulations for PFCs), an open-source tool designed to model hydrogen isotope evolution in ITER PFCs.

• Architecture: HISP acts as a middleware between plasma edge codes (e.g., SOLEDGE3X-EIRENE, SOLPS-ITER) and the hydrogen transport code FESTIM (v2.0).
  • Input: It takes particle and heat flux outputs from plasma codes.
  • Processing: It transforms these 2D/3D fluxes into spatially averaged 1D inputs using a "binning" methodology.
  • Output: It drives FESTIM simulations to calculate inventory evolution over time.
• Binning Strategy:
  • ITER's First Wall (FW) and divertor are partitioned into 97 representative bins (18 FW rows with sub-bins, 44 divertor bins).
  • The model distinguishes between wetted areas (receiving ion and atom fluxes) and shadowed areas (receiving only atom/radiation fluxes) based on magnetic topology and SMITER heat flux profiles.
  • Sub-bins are created for FW rows to account for non-uniform heat loads (high-wetted vs. low-wetted zones).
• Physics Models:
  • Transport: Based on the McNabb and Foster model, solving for mobile and trapped concentrations separately.
  • Materials: Two primary models were used:
    1. Tungsten (W): Uses two trap types with Arrhenius-based diffusion and trapping/detrapping rates.
    2. Boron (B): Models co-deposited boron layers (1 µm to 100 nm thick) with four trap types. The model assumes no fuel exchange between the boron layer and the tungsten substrate due to lack of solubility data.
  • Boundary Conditions: Includes recombination fluxes (DD, TT, DT) and temperature profiles derived from heat flux calculations across the material layers (W, Cu, water cooling).
• Scenarios Tested: Three 14-day operational cycles were simulated:
  • Scenario A ("Do Nothing"): 10 days DT pulses + 4 days STM (no cleaning).
  • Scenario B ("Just Glow"): Includes GDC pulses during STM.
  • Scenario C ("Capability Test"): Includes GDC, low-power DD pulses during operation, and baking during Long Term Maintenance.

3. Key Contributions

• Development of HISP: A novel, open-source workflow that automates the translation of complex plasma edge data into 1D transport simulations, enabling reactor-scale inventory studies.
• Integrated Multi-Physics Approach: Successfully coupled plasma flux data with detailed hydrogen transport physics (including isotope exchange mechanisms) for both Tungsten and Boron materials.
• Quantitative Assessment of Removal Strategies: Provided the first comparative simulation of baking, GDC, and DD pulses specifically for ITER's geometry and material composition.

4. Key Results

• Inventory Baseline: After 10 days of DT operation, the total tritium inventory in the FW and divertor was approximately 35 g.
• Material Dominance: Despite being thin layers, boron co-deposits held nearly 80% of the total tritium inventory (approx. 28 g), while tungsten held the remainder. This is attributed to boron's higher trap density and atomic fraction compared to tungsten.
• Effectiveness of Removal Techniques:
  • Baking: Proven to be the most effective method.
    • Reduced T inventory in Tungsten by ~88% (FW) and ~40% (Divertor).
    • Reduced T inventory in Boron by ~30% (both FW and Divertor).
    • Note: High-energy traps in boron were not fully emptied, suggesting higher temperatures might be needed for complete removal.
  • GDC:
    • Achieved a 23% reduction in Tungsten FW inventory.
    • Achieved only a 7.9% reduction in Boron Divertor inventory.
    • Limitation: GDC ion flux does not reach the divertor strike points (highest heat flux areas), limiting its efficacy in the divertor.
  • Low Power DD Pulses:
    • Achieved ~10% reduction in FW and ~13% in the Divertor.
    • Operates via isotopic exchange (D replacing T).
• Scenario Comparison:
  • While Scenario C (most frequent cleaning) resulted in the lowest final inventory, the differences between scenarios were marginal.
  • Final inventory values varied by <2% in the FW and <10% in the Divertor across all three scenarios.
  • The dominance of baking in the cleaning schedule meant that adding GDC or DD pulses provided diminishing returns in the short term.

5. Significance and Future Work

• Regulatory Compliance: HISP provides a tool to demonstrate compliance with ITER's strict tritium inventory limits and to optimize maintenance schedules.
• Operational Guidance: The study confirms that while baking is the primary removal mechanism, GDC and DD pulses offer supplementary benefits, particularly in the First Wall.
• Limitations & Future Directions:
  • Uncertainty Propagation: The current model does not propagate uncertainties from plasma codes or material parameters, which could affect the marginal differences observed between scenarios.
  • Data Handling: Large file sizes (up to 15 GB per bin) currently hinder data parsing; optimization is needed.
  • Model Refinement: Future work will include multi-dimensional studies, better modeling of the boron-tungsten interface, and validation against experimental data from ITER, JET, and WEST.
  • Optimization: Further research is needed to optimize baking temperatures, particularly for high-energy traps in boron deposits.

In conclusion, HISP successfully demonstrates that while various cleaning techniques exist, baking remains the dominant strategy for tritium removal in ITER, with boron co-deposits acting as the primary reservoir for tritium retention.