Tritium fuel cycle and self-sufficiency - R&D for DEMO and ... Meeting...R&D for DEMO and required extrapolations beyond ITER Christian Day, ... 9th Gaseous Diffusion 10th Molecular

Tritium fuel cycle and self-sufficiency - R&D for DEMO and required extrapolations beyond ITER

Christian Day,

Project Leader of the EUROfusion TFV (Tritium-Matter Injection-Vacuum) Project

15-18 November 2016, KIT

Introduction

EU DEMO fuel cycle single technology developments

Fuel cycle integration aspects

Conclusions

Outline

Chr. Day | IAEA DEMO WS, Karlsruhe | Nov 2016 | Page 2

EU DEMO Power Plant Definition

DEMO Mission Statement:

“The DEMO power plant has to be a representative fusion power station in terms of predictable power production, fuel cycle self-sufficiency and plant performance thereby allowing an extrapolable assessment of the economic viability, safe operation as well as environmental sustainability for future commercial fusion power plants (FPP).”

Hence, DEMO has to:

Be conceived as single step between ITER and a commercial FPP

Produce net electricity (several 100 MWe), safely and reliably

Be tritium self-sufficient and start up another reactor

Demonstrate all technologies for the construction of a commercial FPP Have a representative (extrapolable)

performance:

Lifetime

Cost

Availability and net efficiency

Waste

..Hence, DEMO is not a large ITER, but different.


G. Federici, FED 2016.

Tritium self sufficiency

Demonstration of tritium self-sufficiency is a central element in the fusion roadmap. To demonstrate tritium self-sufficiency successfully, we need to be successful in three aspects at the same time:

TBR

BUF

FCT

Good tritium breeding ratio Breeding blanket / Outer fuel Cycle

High burn-up fraction Particle exhaust at divertor

Efficient fuel cycle technologies to reduce inventory

We need to have sufficient tritium to start and then to provide:


In a simplistic way, one may think to just scale up the ITER fuel cycle.

What we found was that this will end in a number of issues:

…operational complexity, facility size (cost), …

but most of all:

The tritium inventory may act as a SHOWSTOPPER for DEMO:

…in terms of the start-up inventory (may be too high)

…in terms of the regulatory limit (may not be achievable)

…in terms of excessive cycle times and correspondingly too large inertia of the system (tritium plant becoming a very very large chemical plant)

DEMO is a (pre-commercial) power plant, not a physics device

designed for experimental flexibility

DEMO will be designed around one single operational target point

(within an uncertainty margin due to control safety, stability and to meet unknowns, designed for a metal wall right from the start)….

Central DEMO fuel cycle challenge – Inventory reduction


Generic functional fuel cycle scheme

Torus

Tritium Recovery

Bla

nke

t

Isotope Separation

Storage & Delivery

Water Detritiation

D, T DT

Q2 Q2

D,T

Water, He

Impurities

Water, He

Impurities

Q2

Tritium Extraction

Tritium Accountancy

Q2

Helium / Water

T H,(T)

Coolant Purification

Primary Pumping

Rough Pumping

Fuelling & Plasma Control

Q = H,D,T

Tokamak Tritium

Plant

DT

Inner part

Outer part


B. Bornschein, FED 2013.

Generic functional fuel cycle scheme (2)

M. Abdou, FED 2015.


Fuel cycle implementation CFETR

C.A. Chen, Technical Exchange Meeting, Jan 2016

ITER-style Cryopumps

GC, CD, TCAP


Design driver to advance the fuel cycle architecture


M. Abdou, FED 2015.

Target is - to reduce processing times - to increase fuelling efficiency - to increase burn-up fraction

Introduction



Conclusions

Outline


Innovative new fuel cycle concept - FBS

, PEG

PEG

PEG,

Chr. Day, FED 2016.


Derived with a rigorous systems engineering approach

Separation enables DIRECT INTERNAL RECYCLING

Innovative new fuel cycle concept - FBS

, PEG

+ Bypass to the Tokamak to allow easy ramp-up during dwell ´always´steady state

RESIDENCE TIMES

TRITIUM CONTENT

MINIMISED INVENTORY

PEG

PEG,


Rank Technology

1st Pd-Alloy Permeator

2nd Cryogenic Adsorption on

Molecular Sieve

3rd Getter Bed

4th Cryogenic Freezing

Impurity removal (85%)

Rank Technology

1st Membrane Reactor

(catalyst + permeator)

2nd High Temperature Electrolysis

3rd Catalytic Oxidation

Impurity processing (80%)

Rank Technology

1st Thermal cycling absorption process

2nd Plasma Separation Process

3rd Cryo-distillation

4th Gas chromatography

5th Electromagnetic isotope separation

6th Quantum Sieving

7th Pressure Swing Adsorption

8th Laser Isotope Separation

9th Gaseous Diffusion

10th Molecular Laser Isotope Separation

11th Kinetic Isotope Effects

12th Centrifugation

Primary loop protium removal and isotope sep. (75%)

Isotope re-balancing (75%) Rank Technology

1st Plasma Separation Process

2nd Thermal cycling absorption

process

3rd Cryo-distillation

4th Quantum Sieving

5th Laser Isotope Separation

6th Gas chromatography

7th Pressure Swing Adsorption

8th Electromagnetic isotope

separation

9th Gaseous Diffusion

10th Kinetic Isotope Effects

11th Molecular laser isotope

separation

12th Centrifugation

Rank Technology

1st Magnesium hydride catalysed ball

milled

2nd Depleted uranium

3rd Zirconium cobalt

4th Super diamond nanotubes

5th Ammonia borane SBA 15 (mesoporous

silica scaffold)

6th Sodium alanate

7th Lithium Borohydride

8th Activated carbon

9th MOF- 5 (metal organic frameworks)

10th Magnesium Borohydride

Storage (alt. U-Bed) (63%)

Rank Technology

1st Getter Beds

2nd Molecular Sieve Bed

3rd Cryogenic Molecular Sieve Bed

4th Pd-Ag Membrane

5th Water Gas Shift Reaction

6th Cold Trap

Rank Technology

1st Combined electrolysis and catalytic

exchange

2nd Liquid Phase catalytic exchange

3rd Direct electrolysis

4th Vapor phase catalytic exchange

5th Water distillation

Tritium recovery from water coolant (67%)

From the ranking we derive which R&D has to be done with highest priority

Technology ranking results tritium

Tritium recovery from helium coolant (70%)

Rank Technology

1st BIXS and scintillation depending on

detection case (13 cases altogether)

Accountancy (75%) Success of 30 year fusion research


Exhaust Processing R&D – Technology similar to ITER, but operation conditions different: - Significantly higher PEG concentration reduces hydrogen partial pressure - PEG may be activated - Scaling towards higher throughputs unclear

Protium removal and isotope re-balancing R&D – TCAP is not used at ITER but has been advanced mainly in the defense programmes (US, F, China). - Own complementary work may be needed to get full understanding - Scaling towards DEMO throughputs and resulting complexity must be assessed.

Helium coolant detritiation R&D – Need for performance improvement of existing technology in view of the huge flowrates involved.

Water coolant detritiation R&D - Similar technology as for ITER and also applied in heavy water reactors. However, scaling to DEMO seems to have a significant impact on plant cost.

Main R&D headlines tritium

Dynamic control of the loops R&D – Due to the continuous operation of the fuel cycle avoiding the use of intermediate storage wherever possible, gas distribution, control and tritium accountancy become more important.


New requirement: PEG

A variety of plasma enhancement gases are needed for DEMO: For confinement recovery at a metal wall, for radiative seeding, … Different candidates have different activation progeny.


R. Walker, SOFT 2016

Translation into resource loaded R&D programmes

Considering the existing limitation of resources in the TFV project. At any time in the project the choice of projects to be funded can be re-adjusted immediately.


Technology review outcome BB interface

TRITIUM EXTRACTION from the breeder has to convert the breeder outlet (tritiated water, Q2, carrier species) to an input stream to the tritium plant.

60%

60%

50%

Rank Technology

1st Cold trap and adsorption

columns

2nd Continuous catalytic

membrane reactor

3rd Getter bed

Solid Breeder (HCPB)

Rank Technology

1st Permeator against vacuum

2nd Vacuum sieve tray

3rd Gas-liquid contactor

Liquid Breeder (HCLL, WCLL, DCLL)

87% 75% 50%

Cryo-batch concept retained as reference at the moment

membrane

vacuum+ tritium

PbLi in

PbLi out

c1

c2


Technology review outcome matter injection (1)

Fuelling (ELM pacing ?)

Rank Technology

1st Classical pellet injection

2nd Microwave, laser, railgun

3rd Gas puffing, supersonic jets

4th Compact tori

5th Unmagnetized plasma jet

75% < 50% (with ´zeros´)

For gas injection (PEG, gas puffing, incl. massive gas injection) we rely much on ITER technology.

Matter injection has to provide the function of FUELLING and PLASMA CONTROL.

For core fuelling, we go for pellet injection, however have to advance this beyond the ITER operational window.

Can be fueled from the LFS

Large ratio Pellet mass / Plasma mass Large penetration deposition in the core Moderate plasma

pressure limited drift

Small ratio Pellet mass / Plasma mass Shallow penetration deposition at the edge High plasma pressure large drift

HFS injection mandatory

B. Pegourie, 2016



Operational Parameters core ne = 0.9 x 1020 m-3 > nGw

Fusion power

Particle deposition

profile Pellet physics

Pellet parameter

Pellet technology

R&D pellet physics – Develop a self-consistent physics model that translates plasma parameters in engineering requirements (workflow).

R&D pellet guiding tubes – We do have a need to increase launch speed (over what ITER requires) and effective guiding tube systems.

R&D technology demonstration – An EU pellet test bed is needed to develop all engineering features of the DEMO pellet injector: rate, mass, speed

P.T. Lang, FED 2015

B. Plöckl, FED 2015

Curved guide tube ~ 1 km/s

Free flight option (double stage gas gun) ~ 3 km/s




P.T. Lang, 2015.

The (unwanted) contribution of the pellet injection system to the machine throughput (SOL, Vacuum system) is essential.

Nowadays technology:

launch

≈35%....65%

(given by losses

In the guiding system

and SOL

curvature and speed,

deposition depth)

Technology review outcome vacuum (1)

PRIMARY PUMPING has to provide very high pumping speeds, but – different to conventional applications - due to large flowrates, not due to low pressures.

70% 50% 30%

R&D diffusion pump – Develop vapor diffusion pumps for high throughputs and pressures by integration of additional jet stages.

Rank Technology

1st Vapor diffusion pump (continuous)

2nd Metal foil pump (continuous)

3rd Cryocondensation

4th Cryosorption

5th Cascaded cryosorption

6th Continuous cryosorption

7th Warm turbopump

8th Cold turbopump

Primary pumping

R&D metal foil pump for separation and Direct Internal recycling – The metal foil pump has never gone into commercial applications. Develop technology from fundamental physics.

R&D cryopump – Develop multi-stage cryopumps which – to some extent – would allow to implement a separation capability into cryopumping fall-back solution (for risk mitigation).

R&D topic divertor integration – Contribute to an integrated design of the DEMO divertor by physics-based modeling of the particle exhaust to extract the influence of pumping speed (workflow) Chr. Day, FED 2014

Chr. Day, IEEE Trans. Plasma Science 2014.

Pump with separation function


Technology review outcome vacuum (2)

The most challenging requirement for mechanical ROUGH PUMPING for DEMO is the required tritium compatibility at large throughputs (excluding most conventional solutions due to rotary feedhrough issues).

90% 70% 40%

R&D ring pump – Integrate a tritium compatible liquid metal (mercury) into a liquid ring pump.

Rank Technology

1st Liquid ring pump

2nd Roots pump

3rd Scroll pump

4th Rotary vane pump

5th Screw pump

6th Diaphragm pump

Rough pumping

K. Battes, FED 2015 .

R&D wall outgassing – To respond to one of the high-level requirements: Integrated modelling of DEMO pump-down (dwell phase).

T. Giegerich, Fus. Sci. Technol. 2015


Technology choice vacuum (1)

Metal foil pumping is a central part of the DIR concept

Allows continuous gas separation under vacuum, close to the machine

Works only as a pump for atomic hydrogen isotopes

Thermal or non-thermal atomizers for generating atomic hydrogen available

Experimental investigations currently ongoing in a small scale test set-up

Modelling method required for scaling

Classical permeation ~ (p1

½-p2½)

Superpermeation ~ flux1


HERMES test facility @ KIT

B. Peters, SOFT 2016.

Technology choice vacuum (2)

Linear vapour diffusion pumps as simple, reliable and tritium- compatible primary pumps

Customized design for optimal pumping

Using mercury as operating fluid

Liquid ring pumps with mercury as operating fluid

A full scale tritium-compatible pump will be built and utilized at JET DT operation (2018).

THESEUS test facility @ KIT


Demonstration in JET DT campaign in 2018.

T. Giegerich, SOFT 2016. T. Giegerich, FED 2016.

KALPUREX©:

Karlsruhe liquid

metal based

pumping process for

reactor exhaust gases

Direct Internal Recycling Process


T. Giegerich, FED 2014 .

DEMO fuel cycle chosen technologies

, PEG

PEG

Pellet injection

Metal foil pumping

Mercury based vacuum pumping

Membrane reactor

Water formation

CECE

CD

U-bed

TCAP


Dynamic control

Current DEMO fuel cycle architecture


R. Lawless, TRITIUM 2016.

Introduction



Conclusions

Outline


Tritium migration issues ask for joint assessments

T production

~ 360 g/d

T releases

< 0.002 g/d

D. Demange, 2012

The tritium extraction from the blankets is an interface that sets how the tritium handling load is shared between the blankets and the tritium plant. Similarly is teh tritium recoveyr from teh breeder coolant. It consequentially has to be assessed by all stakeholders (system owners) together.


Example CPS interfaces

The amount of tritium permeation from BB to Coolant loop is necessary to define the αCPS (fraction of coolant treated inside CPS).

A correct assessment of the tritium permeation can derive only from an integrated approach.

BB Coolant

CPS

PHTS & BOP

• ηTES

• Anti-permeation

barriers (PRF)

• Oxide layers

• Anti-permeation

barriers (PRF)

• Oxide layers

• ηCPS

• αCPS

T permeation

• Cooling tubes area

• Cooling tubes

thickness

• Tritium inventory in the cooling channels

• Tritium inventory in the coolant (HT and HTO)

T permeation

BB Designers T simulation PHTS & BOP Designers Materials CPS Designers

Safety

• T release

into env.

Fraction of coolant to be treated in CPS

CPS efficiency and coolant chemistry


CPS requirements (1)

Input data

• Allawable T conc. in the

coolant (c0);

• Tritium permeation from BB

to coolant (FT,p);

• CPS efficiency (ηCPS);

• Permeation Reduction

Factor (PRF)

𝐹𝑐0 = 𝐹𝑐𝑖 +𝐹𝑇,𝑝

𝑃𝑅𝐹𝐵𝐵

𝜂𝐶𝑃𝑆 = 𝑐0 − 𝑐𝑢𝑐0

𝐹𝑐𝑖 = (𝐹 − 𝛼𝐹)𝑐0 + 𝛼𝐹𝑐𝑢

1-2

3-4

2-1

𝛼 =

𝐹𝑇,𝑝𝑃𝑅𝐹𝐵𝐵

𝐹𝑐0𝜂𝐶𝑃𝑆

The calculations have been

performed considering

BB SG CPS

F, c0

F, c0

αF, c0 αF, cu

F, ci

FT, p

1

2

3 4 PRF

Coolant fraction to be

treated inside CPS

• c0= 5ppb;

• FT,p in Case#1 and #2

• ηCPS equal to 0.9 and 0.95

• PRF equal to 1, 10 and 100

Parametric Analysis


A. Santucci, SOFT 2016

Amount of coolant to be treated inside the CPS for different CPS efficiencies and PRF values: HCPB

ηCPS PRFBB αCPS αCPS × F,

kg s-1

0.9

1 0.00083333 2.000

10 0.00008333 0.200

100 0.00000833 0.020

0.95

1 0.00078947 1.894737

10 0.00007895 0.189474

100 0.00000789 0.018947

Amount of coolant to be treated inside the CPS for different CPS efficiencies and PRF values: WCLL

ηCPS PRFBB αCPS αCPS × F,

kg s-1

0.9

1 0.00787037 37.77777

10 0.00078704 3.777778

100 0.00007870 0.377778

0.95

1 0.00745614 35.78947

10 0.00074561 3.578947

100 0.00007456 0.357895

• In ITER // HCPB-TBM: 0.00372 kg s-1

G. Piazza, F4E

• In ITER // entire WDS: 0.0166 kg s-1)

CPS requirements (2)

A. Ciampichetti, FED 2010


Philosophy for Implementation of R&D work –

Complement to experiments

A unified fuel cycle simulator is being developed, which integrates the individual system blocks. This will be based on Aspen Custom Modeller, an equation oriented solver platform for dynamic process simulation, we work to do this in collaboration with ITER.

A predictive model is being elaborated for each system block.

The model is tested and deployed if working, or iteratively improved.

The model (and at a later stage the complete simulator) is a perfect tool to explore design space and conduct parametric variations of influential parameters.

Open, commercially available and fully documented chemical engineering plant software, no in-house codes


Fuel Cycle Simulator based on comm. software platform

Numerical implementation cross-check vs ABDOU, 1985


Philosophy for model preparation

The model can be on various levels… fundamental equations (such as diff. balance and conservation equations),

applicable difference equations (such as HTU, NTU, stage concepts,…),

correlation of representative experiments, zeroth order approaches.

…requiring largely different computational efforts…

…and providing largely different understanding for further system optimization.

Once the model is there, it has to be tested. This requires to have a representative test case, which can be literature data or experiments.

Often, we will find that we miss input information. Then, one has to set up additional side-experiments to generate such input information (thermodynamic properties, transport coefficients, kinetics,...) needed to run the main model so that the results are quantitatively representative.

This is the only accepted driver for experiments (not scientific curiosity…), and even this only if it is known that the side-experiment result has a high impact on the model result.


2 4 6 8 10 12 14 16

R [m]

− 10

− 8

− 4

− 2

0

2

4

6

8

10

Z[m

]

Pellet injection engineering

Detached Divertor modeling

Core transport model and burn control

Sub-divertor neutral flow and recycling modelling

Pellet deposition and ablation modelling

Particle exhaust und pumping engineering

SOL loads to exhaust system

Integrated Physics model of the (physics relevant) part of

the fuel cycle

Holistic approach: Physics-Engineering Integration is a must.


DT machine gas throughput

For steady state DEMO operation fuel replenishment due to DT burn is rather small (~2.6 Pa m3/s) and corresponds to the low burn-up fraction (high T inventory)

DT particle throughput due to plasma outflow strongly depends on the tungsten wall outgassing and pumping . Can be roughly estimated as 40 Pa m3/ s and is being fuelled by HFS pellets.

DT replenishment due to He removal in DEMO will very much depend on He enrichment factor in divertor. For limited He concentration in core ≤ 5% one needs to inject about 180 Pa m3/s for enrichment factor value higher than 3%.

LFS GP about 75 Pa m3/s will required for generation of sufficient neutral pressure and low power loading in the divertor (detachment).

Pellet-induced ELMs remove DT particles and are replenished by injected HFS or LFS pellets ~19 Pa m3/s .

N=1x1021/s = molecular gas throughput of 1.7 Pa∙m³/s (referenced to T=273.15 K)

This sums up to minimum 300 Pam³/s DT + ELM pacing gas load + PEG + SOL losses (mainly from pellet injection pumping) Burn-up fraction of maximum 1 %.


DT machine gas throughput with high burn-up fraction

In this case, the DT burn-up fraction can be as much as 10 %, and the DT machine throughput is significantly reduced.

σvnταβ4

σvnταβf

E

E

fuel

burnb

MeVTσvnτE 5.3/)0(12

With (via Lawson-criterion):

and T(0)=26 keV (PROCESS)

for „clean“ wall b→1, ab ~ 1/3÷1/5 , fb ~ 0.5%,

fuel~ 490 Pa-m3/s

for fburn ≥ 10%, (ab ~ 2) tp* ~ 6÷10∙tp →

fuel = burn / fb ≤ 26 Pa-m3/s


Concept to increase ß:

Continuous Exhaust gas re-injection to artificially increase recycling R

H. Zohm, 2015.

Compression due to re-injection He enrichment due to atomic physics

p

pumpburnfuel

VnVv

n

t

4

2

Epp tabtbt

Particle exhaust modelling

Collisionless, x=0.3 Collisional, x=0.3 Collisional, x=1.0

Pressure maps for two

extreme virtual cases:

With and without dome

Calculated with the Direct

Simulation Monte Carlo

(DSMC) code DIVGAS

developed at KIT.

The neutral flow field in the sub-divertor which results from the plasma boundary, the exchange of particles via refluxes, and the vacuum pump capture coefficient, plays a role for the high density scenarios foreseen in DEMO. The EU DEMO divertor development for the first time will be an integrated effort, joining, physicists, material engineers and vacuum engineers. Particle transport to quantify fb is possible.


S. Varoutis, 2016.

Summary and Conclusions

A comprehensive programme is being implemented to advance the

DEMO fuel cycle towards a conceptual design in the next 10 years.

This enterprise is following a system engineering approach to make all

decisions fully traceable and more easily adaptable, if requirements

change.

We propose a new inner fuel cycle architecture to be best fit-to-

purpose, driven by the need to minimise inventory and increase burn-

up fraction, characterized by 3+1 loops.

It is essential to implement an integrated and holistic view on the fuel

cycle.


Documents

Tritium fuel cycle and self-sufficiency - R&D for DEMO and ... Meeting...R&D for DEMO and required extrapolations beyond ITER Christian Day, ... 9th Gaseous Diffusion 10th Molecular