H. D. Pacher1, A. S. Kukushkin2, G. W. Pacher3,

1
Impurity seeding and scaling of edge parameters
in ITER
H. D. Pacher1, A. S. Kukushkin2, G. W. Pacher3,
V. Kotov4, G. Janeschitz5, D. Reiter4, D. Coster6
1INRS-EMT, Varennes, Canada 2ITER Organization,
Cadarache, France 3Hydro-Québec, Varennes,
Canada 4FZ Jülich, Germany 5Forschungszentrum
Karlsruhe, Germany 6Max-Planck IPP, Garching,
Germany
presented at PSI2008 18th Int. Conf. on
Plasma-Surface Interactions Toledo, Spain May 2008
2
Outline
1
1. Results for nonlinear neutral model with
carbon Update scaling (PSI2006, IAEA 2006) 2.
Edge/divertor simulation Impurity-seeded
carbon-free divertor 3. Core simulations Effect
of impurity seeding on ITER operating
diagram Conclusions
3
Edge/Divertor Model
2
B2-Eirene (SOLPS4.3) Now routinely nonlinear
neutral modelneutral-neutral collisionsD2
molecular kineticsParallelized
2 domes
1. Full carbon wall C sputtering phys. const.
Ych 2. Carbon-free wall with neon (wall same as
1. but no C erosion) 3. Variant Full Be wall
with neon (gt with Ne small difference from 2.,
Be concentration small, Be radiation small)
4
Scaling Update
3
Scaling results PSI2006 Key parameter is
normalised pressure
Power
DT pressure at PFR (Throughput)
Factors
Size
is at detachment of either
divertor
Same curve for qpk from JET at 16 MWto DEMO at
500 MW(2006)
Edge density limit Density analogue of from
n scaling
Previous was linear neutral model except for some
points gt Update required.
5
DT flux - Scaling with m and S
4
With nonlinear neutral model Both domes, both S
DT neutral influx to core linear model was
i.e. stronger variation Value at
is 2.4 times that previously Total influx is
still small i.e gas puffing provides little
core fuelling (opacity to neutrals)
6
Helium - Scaling with m and S
5
With nonlinear neutral model Helium slight
difference between domes "bump" at
related to detachment Scaling
linear model was

• Value at is about 1/3 previous
  (linear)
• Helium small, rises less strongly toward lower
  pressures
• helium does not constrain operation unless
  pumping reduced strongly or dome removed

7
Neon seeding without carbon - power
6
With neon seeding Impurity radiation for
similar to but smaller than for C varies little
with (self-consistent carbon is)
But Impurity radiation in inner divertor volume
is smaller than with C
8
Ne - T
7
With neon Temperature at inner target higher (
C chemical erosion, low Ne radiation at lt 10
eV) No additional factor in for
detachment With carbon-free and neon at inner
target compared to Cplasma power
higherradiation lower,total power a bit
higherpower load higher (peaking)
gt peak power load shifts from outer to inner
target (see next)
9
Ne - peak power
8
With neon Peak power shifts to inner target
Only points for which larger load is at divertor
plotted below
Peak power has same scaling but is 30 lower than
with C (but flux expansion and angle are not
same)
10
Ne - nDT and ne
9
The strongest effect of neon is As neon density
increases, gt DT density decreases strongly
Factor 40 broadly consistent with ratio of
ionisation energy 100 for 8ltZlt9 - Less power
available for DT recycling
gt ne decrease 80 of nDT decrease over range
varies little gt explains why neon
radiation varies little with concentration
11
Ne - helium
10
As neon increases to relative to carbon helium
density at separatrix progressively decreases by
2.2 helium neutral influx to core progressively
decreases by 7.5

• Tentatively attribute to
• lower DT and electron densities in divertor
  plasma
• gt lower opacity of inner divertor plasma to
  neutrals
• gt more efficient pumping
• gt lower He densities and fluxes upstream
• Details to be worked out
• (He reduction stronger in inner divertor)

12
Core/edge model
11
Core transport in Astra MMM, stabilised by ExB
and magnetic shear,time-average ELMs, sawteeth,
fitted to JET and AUG, also fits Sugihara
pedestal scaling (EPS2008)
linked to edge via scaling relations from
B2-EIRENE

• profiles, self-consistent pedestal width and
  height
• operating window
• cf e.g IAEA2006, also paper submitted to Nucl.
  Fusion

Because of the opacity of the ITER SOL, core
fuelling controls mostly the core density, gas
puffing controls mostly the edge (including peak
power load via m)
13
Operating diagram
12
Peak power load set given value across
window by varying gas puffing (throughput and m)
At constant alpha particle power has
maximum as n increases gtLow temperature limit
of alpha power from fusion cross-section
14
Operating diagram limits
13
For qpk always within limit, a point lies within
the operational window if for that point
Max. attainable alpha power (roll-over of Pa with
ltnegt
H-L transition
Edge density limit (full detachment)
min Q for ITER mission(5 for ITER)
Available heating power (73MW)
15
Operation at qpk 5 MW/m2 C vs. C-freeNe
14
If very low peak power load required
With carbon Excessive gas puff (up to 300
Pa-m3/s) would be required but density limit
dominates,gtlimiting the actual throughput
usedgtlimiting alpha power
With carbon-free and neon Q is lowerless gas
puff neededgt alpha power is higher
16
Operation at qpk 10 MW/m2 C vs. C-freeNe
15
If moderate peak power load required
The operating window is smaller in both Q and
alpha power with neon Core fuelling a bit
higher at the same alpha power with neon because
of increased core radiation (lower T) Gas
puffing much lower with neon (60 vs 120 Pa-m3/s)
since favourable scaling of peak power in
previous section demands less gas puffing in
addition.
17
Operating diagram summary
16

• Superposition shows
• For Ne relative to C
• low peak poweradvantage with neon in alpha
  power, disadvantage in Qsame throughput at
  limit
• moderate peak power disadvantage both in Q and
  alpha power throughput lower by a factor 2 (but
  lower S would do the same thing)
• edge density limit plays strong role with C,
  less so with neon

18
Conclusions
17
1. Scaling from edge modelling with nonlinear
neutral model updated - DT neutral influx to
core higher than previous but remains small
-gt gas puffing ineffective for core
fuelling - more benign helium scaling toward
lower pressure -gt helium is not strong
constraint for ITER unless pumping strongly
reduced 2. Edge carbon-freeneon relative to
carbon - radiated power similar, peak power
load 30 lower, but with same scaling - peak
power load shifted to inner divertor - helium
density and helium influx to core even lower -
DT density decreases as neon density increases
because less power is available for DT 3.
Operating window carbon-freeneon relative to
carbon - Operating window increases relative
to carbon only when stringent low peak power is
specified, which would require excessive gas
puffing - For moderate specified peak power
loading - neon reduces operating window -
but also reduces throughput (as does reducing
pumping speed, with less impact) gtMutually
consistent modelling of edge and core shows that
for ITER operation carbon-free operation with
neon seeding offers only limited advantage over
carbon operation Carbon-free operation at low
seeded impurity levels needs to be examined
further, as does W