A26792 Text - General Atomics Fusion Group

GA–A26792
ELM PACING BY PELLET INJECTION ON DIII-D
AND EXTRAPOLATION TO ITER
by
L.R. BAYLOR, N. COMMAUX, T.C. JERNIGAN, P.B. PARKS, T.E. EVANS,
T.H. OSBORNE, E.J. STRAIT, M.E. FENSTERMACHER, C.J. LASNIER, R.A. MOYER,
and J.H. YU
JUNE 2010
DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the United States
Government. Neither the United States Government nor any agency thereof, nor any of their
employees, makes any warranty, express or implied, or assumes any legal liability or
responsibility for the accuracy, completeness, or usefulness of any information, apparatus,
product, or process disclosed, or represents that its use would not infringe privately owned rights.
Reference herein to any specific commercial product, process, or service by trade name,
trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any agency thereof. The views
and opinions of authors expressed herein do not necessarily state or reflect those of the United
States Government or any agency thereof.
GA–A26792
ELM PACING BY PELLET INJECTION ON DIII-D
AND EXTRAPOLATION TO ITER
by
L.R. BAYLOR,* N. COMMAUX,* T.C. JERNIGAN,* P.B. PARKS, T.E. EVANS,
T.H. OSBORNE, E.J. STRAIT, M.E. FENSTERMACHER,† C.J. LASNIER,† R.A. MOYER,‡
and J.H. YU‡
This is a preprint of a paper to be presented at the
Thirty-Seventh EPS Conference on Plasma
Physics, June 21-25, 2010, in Dublin, Ireland and to
be published in the Proceedings.
*Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Lawrence Livermore National Laboratory, Livermore, California.
‡
University California-San Diego, La Jolla, California.
†
Work supported by
the U.S. Department of Energy under
DE-AC05-00OR22725, DE-AC05-00OR54698, DE-FG03-95ER54309,
DE-FC02-04ER54698, DE-AC52-07NA27344,
and DE-FG02-07ER54917
GENERAL ATOMICS PROJECT 30200
JUNE 2010
ELM PACING BY PELLET INJECION ON DIII-D AND EXTRAPOLATION TO ITER
L.R. Baylor et al.
ELM pacing by pellet injection on DIII-D
and extrapolation to ITER
L.R. Baylor1, N. Commaux1, T.C. Jernigan1, P.B. Parks2, T.E. Evans2, T.H. Osborne2,
E.J. Strait2, M.E. Fenstermacher3, C.J. Lasnier3, R.A. Moyer4, and J.H. Yu4
1
Oak Ridge National Laboratory, Oak Ridge, TN, USA
2
General Atomics, San Diego, CA, USA
3
Lawrence Livermore National Laboratory, Livermore, CA, USA
4
University of California San Diego, La Jolla, CA, USA
1. Introduction
Deuterium pellet injection experiments have been performed on the DIII-D tokamak to
investigate triggering of edge localized modes (ELMs) in reactor relevant plasma regimes.
Previously, ELMs have been observed to be triggered from fueling pellets injected from all
locations and under all H-mode operating scenarios in DIII-D [1]. Experimental details have
shown that the ELMs are triggered before the pellets reach the top of the H-mode pedestal,
implying that very small shallow penetrating pellets are sufficient to trigger ELMs. Pellet
ELM pacing has been proposed as a method to prevent large ELMs that can damage the
ITER plasma facing components [2]. As part of the experiment on ELM triggering, a
demonstration of pellet ELM pacing has been achieved on DIII-D with a 5x increase in ELM
frequency from the natural ELM frequency.
Experimental details of the pellet ELM pacing
demonstration are reported here.
2. Pellet ELM Triggering
The DIII-D pellet injector [3] was
configured to inject deuterium pellets from
the vertical low field side (V+3) and outside
midplane (LFS) for these experiments as
shown in Fig. 1. The low field side injection
locations were chosen because of the higher
sensitivity of these locations to trigger ELMs
with fueling pellets and lower fueling
efficiency [1]. The smallest available size
Fig. 1. Cross-section view of DIII-D showing
the pellet injection locations.
General Atomics Report GA–A26792
1
ELM PACING BY PELLET INJECION ON DIII-D AND EXTRAPOLATION TO ITER
L.R. Baylor et al.
pellet, 1.8 mm cylinders of equal length and diameter were injected from two barrels of the
injector. One barrel was connected to each of the injection ports. The nominal pellet size
contains 2x1020 atoms (4 mbar-L) of deuterium. All of the pellets injected were observed to
trigger ELMs. A tangential viewing fast camera was available to image the vertically injected
pellets [4]. Divertor Dα emission, divertor infrared (IR) camera, and fast magnetic probes are
all used to diagnose the ELMs in these experiments.
Images from the fast camera of the pellet entering the plasma from the vertical low field
side, as shown in Fig. 2, show the pellets becoming visible from ablation that occurs before
the pellet reaches the separatrix. This is similar to what was observed from pellets dropped
into the plasma in earlier experiments which was attributed to fast ion ablation in the scrapeoff-layer. When the fueling pellet just reaches the separatrix (±1 cm), a single plasma
filament becomes visible just in front of the pellet cloud. This filament is observed to strike
the outer vessel wall within 200 µs of its formation. Additional ejected filaments near the
pellet are then observed to subsequently reach the wall.
Fig. 2. Images from fast camera in Dα light showing a pellet just reaching the separatrix with a filament
becoming visible just in front of the pellet. On the right image, 200 ms later, the filament can be seen to be
striking the wall at the edges of the port. The bright spot below the pellet is a reflection from the wall.
3. Pellet ELM Pacing Demonstration
A demonstration of pacing of ELMs on DIII-D was made by injecting slow
(100-150 m/s)
pellets on the low field side in an ITER shape plasma with low natural
ELM frequency and a normalized β of 1.8 with neutral beam injection heating. Both pellet
injector barrels injected pellets at 7 Hz, alternating pellets between barrels, giving a total
repetition rate of 14 Hz. A comparison of the density evolution and divertor Dα evolution in
two similar discharges from this demonstration is shown in Fig. 3. The density is not
observed to be increased by the shallow penetrating pellets compared to the non pellet
discharge. The plasma energy loss from each ELM in the two discharges determined from
high time resolution equilibrium analysis is shown in Fig. 4. The non-pellet discharge natural
General Atomics Report GA–A26792
2
ELM PACING BY PELLET INJECION ON DIII-D AND EXTRAPOLATION TO ITER
L.R. Baylor et al.
ELM frequency was ~5 Hz with ELM
energy losses up to 85 kJ (>10% of total
stored energy) while the case with pellets
was able to demonstrate ~25 Hz ELMs
with an average ELM energy loss less than
22 kJ (<3% of the total). The resulting
ELM frequency was larger than the pellet
frequency indicating both a direct ELM
trigger by each pellet and an indirect effect
on the overall pedestal stability to ELMs
from the multiple pellets. The energy
confinement time as determined by the
Fig. 3. Temporal evolution of line integral density
and divertor Dα emission in two similar
discharges, one with 14 Hz pellets and other with
no pellets. Pellet injection times are shown with
arrows in the divertor Dα plot.
EFIT equilibrium code shows a modest
~10% lower average confinement time
for the pellet paced ELM case compared
to the non pellet case. The neutral beams
were modulated in these discharges to
maintain a constant β, which leads to a
large
uncertainty
in
the
energy
confinement determination.
Determination of the divertor heat
flux at one toroidal location from the IR
camera data was made during these
discharges [5,6]. The inner divertor peak
heat flux (at the one toroidal location
Fig. 4. The decrease in stored energy for each ELM
in both the non-pellet natural ELM case (blue
circles) and the pellet ELM pacing discharge (red
squares). The period of the pellet injection is shown
by the yellow bar from 2500-3500 ms.
where it is measured) was reduced by an
average of ~50% and the outer divertor by 66% when the pellets were injected as compared
to the non-pellet discharge. The average total energy deposited per ELM to the inner divertor
was reduced by a factor of 3 and to the outer divertor by a factor of 4. The total energy
deposited on the outer divertor is larger than on the inner divertor by a factor of 4. Detailed
analysis of how the heat flux footprint in the divertor changes with pellet injection is
underway.
General Atomics Report GA–A26792
3
ELM PACING BY PELLET INJECION ON DIII-D AND EXTRAPOLATION TO ITER
L.R. Baylor et al.
The rotation of carbon ions in the plasma was measured by a charge exchange
recombination emission diagnostic. The toroidal rotation speed near the top of the pedestal
was observed to decrease from 55 km/s in the non-pellet discharge to 35 km/s when the
pellets were pacing the ELMs. The role of the rotation change in the resulting ELM
frequency is under study.
4. Summary and Discussion
The low field side injected pellets in DIII-D have been found to begin the ELM crash
process within ±1 cm of the pellet crossing the separatrix. This is somewhat shallower than
was observed on AUG [7] with HFS injection. The top of the pedestal in this vertical
injection configuration is >5 cm inside of the separatrix. This implies that the ITER
requirement for low field side injected ELM pacing pellets to reach the top of the pedestal
may be overly conservative. Pellets may need to penetrate just inside the separatrix in ITER
in order to guarantee triggering of an ELM. Future studies with smaller pellets are planned to
elucidate just how far inside the separatrix the pellets need to penetrate. The filaments that are
released during the ELM crash triggered by the pellet are observed to hit the wall locally near
the pellet injection location. While most of the stored energy loss can be accounted for by
extrapolating the heat flux measured in the divertor by the IR camera, it is clear that some
level of heat flux yet to be quantified will impinge on the outer wall. It may be necessary to
provide some high heat flux tolerance capability to the wall surfaces near the pellet ELM
pacing injection location. Further optimization and extension to higher frequency pellet ELM
pacing technique is needed to be able to fully extrapolate this ELM mitigation technique for
application to ITER.
This work was supported by the Oak Ridge National Laboratory managed by UT-Battelle,
LLC for the U.S. Department of Energy under Contract Nos. DE-AC05-00OR22725,
DE-AC05-00OR54698,
DE-FG03-095ER54309,
DE-FC02-04ER54698,
DE-AC52-07NA27344, and DE-FG02-07ER54917.
[1]
[2]
[3]
[4]
[5]
[6]
[7]
L.R. Baylor, et al., Nucl. Fusion 47 1598 (2007).
P.T. Lang, et al., Nucl Fusion 44 665 (2004).
S.K. Combs, et al., J. Vac. Sci. Tech. A 6 1901 (1988).
J. H. Yu, et al., Phys. Plasmas 15 032504 (2008).
D.N. Hill, et al, Rev. Sci. Instrum. 59 1878 (1988).
A. Hermann, J. Nucl. Mater. 337-339, 907 (2005).
G. Kocsis, et al., Nucl. Fusion 47 1166 (2007).
General Atomics Report GA–A26792
4
Download PDF
Similar pages