LBNF Tritium mitigation by scrubbing Jim Hylen TSD topical meeting - PowerPoint PPT Presentation

LBNF Tritium mitigation by scrubbing Jim Hylen TSD topical meeting 21 November 2019

DAC (Defined Air Concentration) • DAC is amount of radiation in air that: Would gives 5 rem dose to workers in 2000 hr work year. = 2.5 mrem/hr, = 100 mrem in 40 hr work week. • For tritium in form of HTO, DAC is 20 pCi/cc of air. (Humidity) • At 25 C, 100% RH, it translates to about 852,000 pCi/ml of water At FNAL, we don’t let water get tritiated much beyond 800,000 pCi/ml, so that leaks/spills can’t get the air above a DAC. • Work restrictions required if Tritium in air > 10% DAC Note we are trying to design LBNF to stay below 10% DAC 2 Jim Hylen | LBNF Tritium mitigation 11/21/2019

During NuMI operation, MI-65 ventilation AHU condensate up to 1,720 pCi/ml, AHU1, AHU2, AHU3 fairly similar Evaporator ~ 500,000 pCi/ml 600 cfm of water in air Hi-Bay AHU 3 AHU 1 AHU 2 EAV2 Evap 6,000 cfm stair shaft Absorber AHUs release Condensate condensate to sewer Pile Shield Target Hall Dehum door Target Pile 3 Jim Hylen | LBNF Tritium mitigation 11/21/2019

9/11/2018 short circuit to AHU during access at NuMI Target pile open, target hall shield door open, pile dehumidifier off worst condition for exposure Hi-Bay AHU 3 AHU 1 AHU 2 EAV2 Evap 24,300 pCi/ml of water in air Absorber Condensate stair shaft 4% of DAC in 41,500 pCi/ml 950 / 819 / 17,500 pCi/ml target hall of water in air condensate to sewer Pile Shield Target Hall Dehum door 66,600 pCi/ml Target Pile of water in air 4 Jim Hylen | LBNF Tritium mitigation 11/21/2019

Tritium concentration in dehumidifier water, evaporated to air 1) Large jump in 700 kW 700 kW tritium release ops ops when Duratek steel shielding 540 kW got to ~ 100 C ops (not smooth exponential in diffusion versus temperature) Yearly shutdown 2) Lower emissions Yearly shutdown Yearly shutdown Yearly shutdown during beam-off No beam, 300 kW upgrade ops for NOVA 5 Jim Hylen | LBNF Tritium mitigation 11/21/2019

Designing for LBNF access Higher beam power -> more tritium • NuMI 700 kW beam power • LBNF eventually 2,400 kW beam power Potential to trap tritium until a shutdown, then release in extended puff • NuMI leaky pile with humid air continually getting rid of tritium • LBNF sealed pile with dry Nitrogen fill could trap tritium until open pile for access when humid air gets in 6 Jim Hylen | LBNF Tritium mitigation 11/21/2019

GH local exhaust local exhaust The work wearing airline mask and anorak suit is hot and extremely tough. 2-hour is the upper limit for one work. airline mask anorak suit air supply line NBI- 2019 “Muon Target Replacement” by N. Kawamura (JPARC)

Mitigation strategies being implemented for LBNF • Remove most tritium during beam operations by scrubbing with humidity (NuMI experience) • Operate inner layer of steel (the cooling panels) at higher temperature (~100 C) during operation, then cooler (room temperature) during access to encourage higher/lower release during running/access. (NuMI experience) • During access, increase air ventilation rate to dilute the HTO – 1,091 cfm -> 22,300 cfm of air through target hall • Draw the air (to the extent practical) away from workers to a release stack – Ventilation once-through, from target hall down into target pile, then up stack 8 Jim Hylen | LBNF Tritium mitigation LBNF 11/21/2019

LBNF: scrub tritium out of target pile with humidity during beam Goal: Don’t build up large amount of tritium, which would come out in puff when opening up Target Hall Require addition of mist injection of humidity Dehumidifier for N 2 in vessel N 2 gas was already in design Steel shielding releasing tritium Plan to remove most of the tritium released from the shielding during beam running the same way we do for NuMI – pick it up with water vapor to HTO, and dehumidify. Calculations indicate parameters shown will Comparison NuMI LBNF • maintain the tritium transport path Target pile gas air N 2 • keep built up tritium to reasonable levels H 2 O in gas (ppt) 5 1 We are also optimizing ventilation to protect Condensate (gallon/day) 170 17 workers during access 9 11/21/2019 Jim Hylen | LBNF Tritium mitigation LBNF

During Target Pile access Ventilation to help mitigate high Tritium release in target pile vessel • Pull tritium away from workers in nitrogen vessel • Use existing filter to pick up contamination before exhaust Access target pile and gas handling just one at a time 22,300 cfm Fresh air for workers 22,300 cfm air Target hall Gas handling room Target pile gas system Target pile vessel Nitrogen vessel Steel shielding chase Pre-target OPEN Filter chiller dehum. fan CLOSED OPEN Exhaust to stack 10 Jim Hylen | LBNF Tritium mitigation LBNF 11/21/2019

During Target Pile Gas Handling access Ventilation modification stay away from high Tritium release in target pile vessel • CLOSE off main tritium reservoir (target pile shielding) 0 cfm • Turn off gas handling fan, use some small local ventilation 22,300 cfm from absorber Access target pile and gas handling just one at a time fresh air 22,300 cfm Gas handling room Target hall exhaust Target pile gas system Target pile vessel Nitrogen vessel Steel shielding OFF or chase LOW Pre-target CLOSED Filter chiller dehum. fan CLOSED OPEN Exhaust to stack 11 Jim Hylen | LBNF Tritium mitigation LBNF 11/21/2019

For the tritium scrubbing with humidity Evaporter & stack to air Have set up a framework for tritium transport, Dehumidifier and done 1 st pass calculations: • Model elements – follow the Tritium path: Humidity In N2 gas - Tritium production in steel shielding (MARS MC) - Diffusion rate of tritium through steel Surface water - Steel to surface water layer tritium transfer • (surface water is few molecules thick) Steel - Surface layer tritiated water to humidity exchange - Humidity turnover rate (injection and dehumidification) The model calculations indicate the parameters shown earlier are sufficient to: • Maintain the surface physical-water layer that underpins the transfer calculations • Maintain a high gradient from steel to surface water layer for tritium transport • Keep the amount of tritium stored in the target pile humidity to reasonably low level The model will be further refined and checked, and some prototyping may be done 12 11/21/2019 Jim Hylen | LBNF Tritium mitigation LBNF

First model iteration – assume equilibrium, locate possible rate limiting steps The steel blocks need thin layer of water – (OH layer possible as base) • Nishikawa et al, Journal of Nuclear Materials 277 (2000) 99- 105 “Tritium trapping capacity on metal surface ” subdivides surface water as – Structural water – Chemically bound water – Physical water – most easily exchanged with gas • The formula they provide for physical water is calculated from partial pressure of H 2 O, temperature, and a constant that depends on the metal surface – In the range of 20 C to 100 C for their materials this is around a few E-4 mole/m 2 I put 1e-4 mole/m 2 in the model to check if this essentially covers the surface, or we end – up with partially dry surface -> looks fully covered to me – Note: used input H 2 O pressure 101 Pa (i.e. 1 atm at 1000 ppm H 2 O) – Note: very generic rough; our blue blocks and water cooling panels may vary • To get order of magnitude estimate of tritium stored in the water on surface of steel, took WAG of 5000 m 2 of steel surface, 3e-4 mole/m 2 of water: get ~ 1 mCi 13 Jim Hylen | LBNF Tritium mitigation 11/21/2019

14 Jim Hylen | LBNF Tritium mitigation 11/21/2019

First model iteration – assume equilibrium, locate possible rate limiting steps Molecular exchange between humidity and surface layer • I used our NuMI experience with tritium leaching out of the air when it went down the decay pipe passageway to WAG how fast the humidity exchanges with a water layer. The time constant for NuMI was << 1 hr. I cross-checked this with some literature where people were doping tritium onto samples, and removing tritium from samples. It needs to be documented in a more refined manor, but this exchange rate appears to be fast enough to not be rate-limiting in this version of the calculation. Recall time scale of the diffusion in steel to water layer is days. 15 Jim Hylen | LBNF Tritium mitigation 11/21/2019

First model iteration – assume equilibrium, locate possible rate limiting steps Exchange rate between surface water and steel • My impression from those same studies of tritium is the transport between water layer and near-surface steel is fairly fast, and should not be rate-limiting, but need better documentation. • Note there are barriers, like an oxide layer on stainless steel, that can significantly slow down the transport. A cautionary statement in fusion research is “barrier layers tend to break down in radiation environment”. • The T in the steel can basically trade places with an H in the water layer • Is the layer of black iron-oxide at T2K on the vessel walls why there seems to be significantly less release of tritium at T2K than NuMI? 16 Jim Hylen | LBNF Tritium mitigation 11/21/2019