UCRL-PROC-200898
A Fusion Chamber Design with a
Liquid First Wall and Divertor
R.E. Nygren, D.K. Sze, B.E. Nelson, P.J. Fogarty, C.
Eberle, T.D. Rognlien, M.E. Rensink, S. Smolentsev,
M.Z. Youssef, M.E. Sawan, B.J. Merrill, R. Majeski
November 12, 2003
20th Symposium on fusion Engineering
San Diego, CA, United States
October 14, 2003 through October 17, 2003
Disclaimer
This document was prepared as an account of work sponsored by an agency of the United States
Government. Neither the United States Government nor the University of California 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 the University of California. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States Government or the University of California,
and shall not be used for advertising or product endorsement purposes.
1
Abstract: The APEX study is investigating the use of free
flowing liquid surfaces to form the inner surface of the
chamber around a fusion plasma. We present a design for
the chamber of a 3840MW fusion reactor based on the
configuration for the chamber and magnets from ARIES-
RS but with a fast flowing molten salt of mixed Be, Li and
Na fluorides for the first wall and divertor and molten salt
blanket with a ferritic steel structure. Our design analysis
includes strong radiation from the core and edge plasma,
(liquid) MHD effects on the weakly conducting molten salt,
a recycling first wall stream that enables a high efficiency
thermal conversion, and evaluations of breeding,
neutronics, tritium recovery and safety.
.
I. INTRODUCTION
This paper briefly summarizes a design for the chamber
of a magnetic fusion reactor with flowing liquid walls
facing the plasma. While removing surface heat, the
continuously-renewed liquid surface is unaffected by the
radiation damage and thermal stresses that limit the
performance of solid chamber walls. A companion paper
summarizes the divertor design[1] and a longer paper[2]
provides more detail and includes extensive references
and discussions of R&D issues. The design effort, part
of the Advanced Power Extraction (APEX) study[3],
utilizes work on plasma interactions with liquid surfaces
performed in the Advanced Limiter-Divertor Plasma
Facing Systems (ALPS) study[4]. The ideas for fusion
reactors with liquid metals and even liquids with free
surfaces are not all new[5-23], but new ideas have been
developed in APEX and the depth of the underlying
science and engineering exceeds previous work.
Chamber technology serves two of three fundamental
functions of fusion energy systems. The first is confining
the plasma itself. The second and third are (1) breeding
of sufficient tritium to fuel a self-sustained reactor, and
(2) practical, reliable, safe, economical power extraction.
Economics pushes fusion toward high power density and
high-temperature coolants; these set requirements for
.
*Sandia is a multi-program laboratory operated by Sandia
Corporation, a Lockheed Martin Company, for the United States
Department of Energy under Contract DE-AC04-94AL85000.
tritium breeding and heat removal.[3] Three principal
features of the fusion chamber are the (1) first wall and
divertor, (2) the blanket, and (3) the shield (Fig. 1). We
also include supporting systems such as those for tritium
processing and heat exchange. We have adapted the
basic magnetic configuration from an existing design
study for a 2170 MW D/T fusion power plant called
ARIES-RS[24,25] with 16 toroidal field coils, a major
radius of 5.5 m and aspect ratio of 4, and incorporated a
liquid surface first wall and divertor and a liquid blanket.
We have explored mechanical designs for several
concepts and each has included detailed CAD renderings
and several innovative features.[2] In 1999-2000, we
studied a design with the molten salt Flibe as the
working liquid, and, in 2000-2001, designs with Li and
with Sn-Li. Excessive surface temperature (vaporization
of F) limited the Flibe design, and poor thermal
efficiency due to the relatively low temperature for
evaporation of Li limited the Li design.[26] In 2001 we
began designs with a Sn first wall and Pb-Li blanket.
The analysis of plasma surface interactions gave a fairly
good operating temperature range with the surface
temperature limit of Sn being 810-840°C for the FW and
1630°C (1480°C for Ga) for the divertor. Simple
correlations for the MHD forces and the assumption of
an insulating wall showed that ~20mm-thick flow
streams launched at 10 m/s would flow down and adhere
to concave substrate. The difficulty of modeling the
magneto-hydrodynamic (MHD) effects that dominate
the fluid flow precluded more detailed analyses, and this
area is being actively investigated through both
modeling and experimental work in APEX.[27]
We refer readers to another paper[2] for descriptions of
many innovative features. For example, in utilizing a
Sn-Li mixture, the lower activity of the Li in the mixture
and lower evaporation rate raised the allowable surface
temperature of the first wall from 380°C for Li to 590°C
for 0.8Sn-0.2Li[28,29], and segregation of Li to the
surface (driven by a reduction in the surface tension)
was also identified as an issue and possible benefit.[30]
A Fusion Chamber Design with a Liquid First Wall and Divertor
R.E. Nygren, D.K. Sze
a
, B.E. Nelson
b
, P.J. Fogarty
b
, C. Eberle
b
,T.D. Rognlien
c
, M.E. Rensink
c
,
S. Smolentsev
d
, M.Z. Youssef
d
, M. E. Sawan
e
, B.J. Merrill
f
and R. Majeski
g
Sandia National Laboratories
*
, PO Box 5800, Albuquerque, NM 87185
a
University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0417
b
Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831
c
Lawrence Livermore National Laboratory, 7000 East Ave., Livermore, CA 94550-9234
d
University of California, Los Angeles, MAE Dept. Box 951597, Los Angeles, CA 90095-1597
e
University of Wisconsin, Madison; Madison, WI 53706
f
Idaho National Engineering and Environmental Laboratory, P.O. Box 1625, Idaho Falls, ID 83415
g
Princeton Plasma Physics Laboratory, P.O. Box 451, Princeton, NJ 08543-0451
UCRL-PROC-200898
2
Inboard
Blanket
Module
3/Sector
Shield
Module
Outboard
Blanket
Module
5/Sector
Heating &
Current Drive
Modules
substrate for flowing liquid first wall
blanket back wall
nozzle
module
outboard
first wall
flow
in-
board
first
wall
flow
divertor
module
deflector
shield liner & drain
Figure 1. (left). Chamber components: blanket and shield modules (top), first wall and
divertor flow paths and modules (bottom).
Figure 2. (above) UEDGE “maps” of fluorine radiation. See text.
Figure 3. (below) Flow streams and bulk Flinabe temperatures (right) and cross-section of
outboard blanket module (left).
3
Other examples include modifiying the flow substrate to
improve flow around penetrations, flexible SiC bags that
contained the liquid blanket, and "self shielding" nozzles
invented by authors Nelson and Fogarty.
In our designs, modeling of the plasma edge defines the
heat loads, and thermal-hydraulic analyses show that the
heat removal is adequate. Evaluation of thermal
transport shows an operating point with reasonable
thermal efficiency and neutronic analyses show adequate
tritium breeding. Evaluations of safety indicate
manageable approaches to the issues. Since our focus is
on chamber technology, we do not develop descriptions
of the magnets, balance of plant, reactor hall, etc.
In FY2002 we began our design with the molten salt,
Flinabe, described later. It has similarities to Flibe but a
lower melting point. We began with some skepticism
since molten salts have poor thermal conductivity and
good thermal conductivity would seem to be an a priori
requirement for any first wall and divertor material. In
this regard, our APEX/ARIES design with Flinabe is a
startling and pleasing result.
II. CHAMBER CONFIGURATION
In our adaptation of ARIES-RS, we increased the power
density by specifying a fusion power of 3840 MW to
define the heat loads for the first wall and divertor, so
that a total power of 909 MW (767 MW alpha power,
142 MW auxiliary power) will be transmitted as surface
heat loads to the first wall and divertor (Table 1).
TABLE I
POWER BALANCE PARAMETERS FOR APEX CHAMBER SYSTEM
First wall area
434
m
2
Average neutron wall loading
7.08
MW/m
2
Average surface heat flux
1.76
MW/m
2
Fusion power
3840
MW
Alpha power
767
MW
Auxiliary power to plasma
142
MW
Power to the first wall
765
MW
FW nuclear heating
400
MW
Total power removed by FW
1165
MW
Power to the divertor
144
MW
FW surface heat flux, average
1.76
MW/m
2
FW surface heat flux, maximum
3.5
MW/m
2
Neutron power
3073
MW
Neutron wall loading, average
7
MW/m
2
Neutron wall loading, maximum
10
MW/m
2
Blanket energy multiplication
1.06
Blanket thermal power
3257
MW
Total thermal power
4024
MW
Total power (thermal + aux.)
4166
MW
In high power density devices, a large fraction of the
alpha power must be converted into radiation to make
the heat load on the divertor manageable. While large
radiative cooling can mitigates the peak heat loads, it
also tends to decrease the thermal isolation and stability
of the core plasma. In our design, strong radiation from
the core plus strong radiation from the divertor region is
used to reduce the peak heat load in the divertor and
balance the power loads of the divertor and first wall.
The power to the first wall of 765 MW is 84% of the
total particle power of 909 MW while the balance of
power to the divertor (144 MW) is only 16%.
The recent plasma transport modeling by Rognlien and
Rensink, finds a stable operating window for a highly
radiating edge plasma including steady state modeling
solutions in which about 95% of the power coming into
the scrape-off layer is radiated near the X-point for alpha
powers in the range of 300-360 MW. This work is
described briefly in our companion paper[1] and
elsewhere[2,31]. Figure 2 shows a sample result for a
series of solutions at progressively higher values of P
c
,
power convected from the plasma core into the edge
plasma. The plots are maps of flourine radiation in the
lower portion of the chamber cross section with strong
impurity line-radiation concentrated near the X-point
and below. The set of four cases shows a progression
from a case that is not stable (MARFE), through two
cases (300 and 400 MW) with stable operation, to the
440 MW case that is unstable and “burns through” onto
the target. The 300 MW case had a peak heat flux
(particle + radiation) at the divertor plate of only ~8
MW/m
2
and a peak heat flux to the first wall of ~2
MW/m
2
for a density at the edge of the core
of ~1.5x10
20
m
-3
, H (D/T) throughput of ~3.1x10
23
particles/s with
divertor plates orthogonal to magnetic flux surfaces and
H pumped at private flux surface for stability. The
Fluorine density at core boundary varies poloidally over
the range 3.7-7.3x10
17
m
-3
(0.24% - 0.49% of hydrogen;
1% is the limit based on core radiation loss). This work
supports the conclusion that it is reasonable to
investigate designs based upon relatively modest peak
heat fluxes (8-10 MW/m
2
) in the divertor.
III. FLOW PATHS AND HEAT TRANSFER
Fig. 3 shows the flow streams and bulk temperatures.
The inboard and outboard first wall streams flow from a
set of nozzles at the top of the chamber that inject the
Flinabe onto the front surfaces of the blanket modules.
One set of nozzles feeds the inner first wall, another set
feeds the outer first wall. The overlapping streams from
the “self-shielding” nozzles prevent any line-of-sight
from the plasma directly to a nozzle.[2]
The thickness of the first wall flow at midplane is 23
mm and the flow speed is 10 m/s. The first wall flow
length is ~6.5 m. The Flinabe starts at 402°C and rises
to 420°C at the bottom of the first wall. The penetration
of the surface heat and nuclear heating contribute about
equally to the rise in bulk temperature of ~18ºC. The
surface temperature, which reaches 509°C at the bottom
of the first wall, was calculated by Smolentsev using a
