ARIES Program
Public Information Site


ARIES-CS Project Meeting Minutes

16 September 2004

Madison, Wisconsin

Documented by L. Waganer

Organization ARIES Compact Stellarator
Boeing Waganer
Columbia Univ.  
GT Yoda
NYU Garabedian
ORNL Lyon, Snead, Zinkle
UCSD Najmabadi, Raffray, Wang (Xueren)
UW-Mad Blanchard, El-Guebaly, Sawan, Sviatoslavsky (Greg), Sviatoslavsky (Igor),Tautges (Tim), Wang (Mengkuo)

Ref: Agenda and Presentation Links: Project Meeting


Welcome - Laila El-Guebaly welcomed the ARIES team to the Monona Terrace at Madison. She did an excellent job of securing a room at the Terrace following the 16th TOFE meeting.

Status of ARIES Program - Farrokh Najmabadi told the group he wanted each group to have a “Scope of Work for 2005” presented to him by the November meeting so he can have a coherent project scope for 2005 (and what is remaining for 2006). He informed the group that the ARIES-IFE papers that constitute the final reporting would be published in the November Fusion Science and Technology Journal.

Next Meeting/Conference Call - Farrokh Najmabadi affirmed that the next ARIES meeting will be held at the University of California, San Diego on November 4 (Thursday) and half a day November 5 (Friday).

Compact Stellarator Reactor Engineering Assessment

Update of the ARIES-CS Maintenance ApproachCS Cross Section Xueren Wang illustrated the improved CAD definition and visualization of the ARIES-CS 3-field period power core configuration with the external vacuum vessel arrangement and field period replacement maintenance. He highlighted the support physical arrangement between warm and cryogenic elements. All heat transfer plumbing is routed through the bottom of the power core for better access and minimal translation displacement relative to supports. Xueren presented a higher fidelity mid-plane cross-sectional view to illustrate the complex geometries of the first walls, blankets, shields, and vacuum vessel. This view shows the position of the localized shielding added to the outside of the blanket.

Xueren also updated the power core arrangement with internal vacuum vessel and modular replacement maintenance. The vacuum vessel is now conformal to the outside of the shield except for the access port region. A revised supporting tube and bucking cylinder is presented.

His future work will be to further define the heat transfer plumbing and supporting structures. He will also investigate approaches to better contain the contamination during maintenance actions. It was suggested a view into the port might be beneficial.

Likely Blanket Structural and Heat Transfer Materials – Steve Zinkle reported on the US fusion community efforts to concentrate research and development on structural and heat transfer materials that will likely be used in the nearer term ITER Test Blanket efforts. He was not downgrading the long-range potential of other materials, but just identifying those materials with sufficient and attractive documented properties for use in the Test Blanket Program. Molten salts (e.g. Flibe or Flinabe) do not have sufficiently documented data for near term use. This decision is reinforced because the Generation IV fission designs use molten salts in only one, lowest priority design approach. Fusion is not planning to fund any molten salt development at present beyond the modest activities within the US-Japan Jupiter-II collaboration. Liquid metals are being proposed for two Gen IV fast breeder reactors. Sodium is used in the fast reactor and lead (or lead-bismuth) in the lead-cooled reactor. The remaining proposed Gen IV reactor concepts use either helium or super-critical water as heat transfer media.

LiPb has been chosen as the heat transfer media and ferritic steel (maybe ODS FS) as the structural material in current leading candidate module for the US ITER Test Blanket program (utilizing SiC flow inserts). A flowing LiPb/SiC test loop will be required to verify the material performance and compatibility throughout the operating temperature range (up to 1100°C). The FS will probably have a corrosion barrier of SiC, if the temperature exceeds 450°C. Some tests of LiPb in silicon carbide containers have been conducted, but more extensive tests are planned.

A lithium-cooled blanket module is also being considered, to be led by the Russians. Japan, South Korea, and the US may be interested in participating. The current proposed reference Li-cooled module uses vanadium alloys. In one option under consideration, the structural material would be FS operating at lower temperatures (up to ~ 550°C) with SiC inserts. Some testing has been accomplished at lower temperatures NASA has been testing a related space power application that uses lithium coolant in Nb-1%Zr structures.

Laila asked Steve to recommend an upper temperature limit for FS during LOCA/LOFA events. For a one-hour duration of the high temperature, Steve suggests a limit of 740°C for FS and 700°C for ODS FS.

Preliminary LOCA Analysis for LiPb/FS System – Jake Blanchard presented his preliminary LOCA and LOFA analyses of the LiPb/FS blanket system for the Compact Stellarator. He assumed the blanket has only a minimal physical contact with the vacuum vessel. The outer surface of the vacuum vessel represented an adiabatic boundary as a conservative assumption. He asked how long to leave the plasma on – the consensus was that the plasma would extinguish in a matter of seconds if the coolants entered the plasma chamber. Laila El-Guebaly informed Jake that the afterheat would be higher than previously thought and she would provide revised values. With complete loss of helium and water (LOCA), but the LiPb contained and not flowing in the power core (LOFA), Jake thought the FW temperature would quickly rise to 1250°C and then slowly continue to climb due to LiPb afterheat. Here, the FS decay heat has been used for the LiPb as well. Laila El-Guebaly informed Jake that the LiPb decay heat would be different and she will provide revised values for LiPb.

For the condition of a LOFA for the LiPb and water except for natural convection, the first wall temperature peaks out around 1300°C in around 0.5 days. A reasonable emissivity variation for the gap surfaces results in an approximate 200°C variation. If the LiPb and water are drained (LOCA for both coolants), the first wall temperature peaks around 875°C, cools slightly, and then continues to climb slowly due to afterheat in the FS. If the LiPb is drained (LOCA) and the VV water is retained but not flowing (LOFA), the first wall temperature peaks around 875°C and then stabilizes around 760°C. Jake will redo the analyses with the corrected LiPb afterheat values and the assumption that the plasma is extinguished within seconds of the accident initiation.

Power Core Options: Decision Factors and Selection for Phase II - René Raffray reviewed the engineering activities leading to a selection of power core elements for next year’s Phase II design optimization effort for the reactor core configuration and maintenance. He summarized five blanket concepts and the two maintenance approaches (field-period and port maintenance) to be considered. He noted the improvements in the designs since the last meeting. He recommended we retain both maintenance approaches in Phase II. He showed the major features and expected performance for the five blanket approaches. It was recommended the blanket design options for Phase II include (1) a good performance, lower development risk, dual-coolant liquid metal and FS structure blanket option and (2) a higher performance, higher development risk, self-cooled LiPb with SiC/SiC structure blanket option.

He showed a qualitative/quantitative, weighted trade study Les Waganer suggested for the assessment and selection of one or more blankets for the next phase. René then qualitatively discussed the pros and cons of each blanket concept without out weighing the relative merits of each and every metric. The ARIES team agreed that a documented trade study should be used to identify and document the important metrics, the relative value of the metrics to the selection process, and expected performance of each metric for each concept. Les will work with René on this trade study process and the metrics.

Summary of Power Core Materials and Fabrication Unit Costs – Les Waganer noted that after the selection of the power core concepts, the power core will be designed and integrated. The concept will then be evaluated with Jim Lyon’s systems code. Les developed the material and fabrication unit costs for the LiPb/FS/He blanket with the internal vacuum vessel concept, based on Laila’s radial build.

Material and fabrication costs were developed for the blanket, shield, and vacuum vessel on a dollars/unit cost basis, assuming 2004 dollars, 10th of a kind fabrication, and a complexity factor based on fabrication approach. For the considered blanket approach, representative, large-quantity unit costs for ferritic steel, tungsten carbide, and lead (for LiPb) were provided.

The likely cost for enriched lithium (high percentages of Li6) cannot be determined or forecast at this time. The US Government was the primary producer and user of Li6 until 1963. Since that time, there has been little commercial usage of Li6. A placeholder price for 90% enriched lithium was provided to conduct the trial costing exercise. However a more reasonable and likely cost must be developed. Laila will provide Les a cost estimate of enriched lithium from UWTOR-M, Report UWFDM-550, dated October 1982. Les will develop the remaining materials for the remaining core elements.

[Update: Les and Laila found an error in Les’ method of calculating the fraction of lithium in LiPb as it should have been atom%, not weight%. Thus the quoted costs in the presentation for enriched LiPb were excessive. The presentation has been updated and is available at the ARIES Web Site.]

Neutron Wall Loading Profile Using CAD/MCNP Interface – Mengkuo Wang repeated a few slides from our last meeting to illustrate how the plasma boundary of the 3-FP configuration is subdivided horizontally into bins for accumulating the particles. He reminded us that the results at the FW differed by approximately 3% due to the 5 cm scrape off layer. He has now modified his interface code to sweep (subdivide) the perimeter of the plasma cross-section into discrete bins. These bins can be combined with the horizontal sweep to create a set of area patches that describe and subdivide the stellarator plasma surface. He has created nine cross-sections that, when combined with symmetry, fully describe the stellarator plasma surface. He has then generated neutron wall loads (NWL) at these nine cross-sections. The peak NWL is at the outboard surface. The peak to average ratio is 1.6.

The results are much higher fidelity than any other existing method. However, the computational needs are 10-30 times greater. His future work is to investigate different methods to improve the computational time.

Compact Stellarator Reactor Physics Basis

Assessment of Quasi-Helically Symmetric Configurations as Candidates for Compact Stellarator Reactors – Long Po Ku assess the ability of a quasi-helically (QH) symmetric stellarator as a configuration for an attractive reactor basis, preferably with an aspect ratio less than 6 and still have good symmetry. HSX has an aspect ratio of 8 with 4 field periods.

3-Field Period Plasma SurfacesLong Po showed a 3-field period configuration with an aspect ratio of 6 (3H4), shown to the left. It has positive shear in the rotational transform, a magnetic well of ~1%, and an effective ripple of less than 0.35%. However the plasma is very tall and thin with a sharp cusp in one area. He also further optimized the configuration (3H5), that yielded lower alpha loss (<0.5%). But this was achieved at the expense of even thinner plasma shapes. The sharp cusp has been softened. Both configurations have high iota, low shear, and low well depth, but are mostly Mercier unstable. By increasing the magnetic well depth to ~ 4%, the configurations are stable to Mercier for beta values >4%.

3-FP QH Configuration Future work is focused on improving the QH configurations to improve the plasma shape, lower aspect ratios, and retaining good QH properties. Long Po mentioned a 3-field period case, shown at the right, with an aspect ratio of 4.5 that he is assessing and will examine before the November meeting.

A Look at Another Prompt Alpha Loss Approach – Paul Garabedian explained his new-found appreciation of the importance of having plasma and coil configurations that will not have excessive prompt alpha loss. He has met with T.K. Mau and Long Po Ku to discuss methods and optimization approaches to refine his models and codes to produce lower alpha loss configurations. He has looked at the quasi-axisymmetric (QAS) configuration to reduce the alpha loss to around 10%. He needs a free boundary code at the first wall surface, but is using the last closed magnet surface (LCMS). He would like to get a new code to define divertor parameters.

Compact Stellarator Reactor Integrated Systems Assessment

Progress on Systems Code and Revision of Previous Results – Jim Lyon summarized the purpose and status of the ARIES Compact Stellarator Systems Code. It is derived from the prior SPPS stellarator systems code. He has completed the coil geometry and updated physics and engineering modeling. He is working on the modeling of impurities and plasma edge and improved modeling of the power core engineering. He will be gathering and implementing the costing and maintenance algorithms in the near future. He explained how the system code is organized and its optimization logic. He has updated the input parameters for four reactor configurations with the NCSX-2 just being added.

Key Configuration Properties
Plasma aspect ratio, Ap = <R>/<a>
Wall (plasma) surface area/<R>2
Min. plasma-coil separation ratio, <R>/Dmin
Min. coil-coil separation ratio, <R>/(c-c)min
Total coil length/<R>
Bmax,coil/<Baxis> for 0.4-m x 0.4-m coil pack

Jim reported that in his 1-D POPCON calculations, he found and corrected an error in the helium calculation, which affected the nDT/ne and Pfusion. This corrected the concern that tHe/te is too low. He is also updating the ISS-95 confinement scaling in the code.

Updated results from the code indicate the NCSX-1 values are limited to a neutron wall load of 5MW/m2. He presented the main parameters from the 0-D code for the four reactor configurations. He recommended the NCSX-1 for more detailed study.

Jim presented the key features of the LHD power plant study being conducted by Tom Dolan. It is based on the LHD plasma and coil configurations, with port maintenance. Tom is evaluating three blanket options:

  • RAFS and Flibe
  • V and lithium
  • SiC and LiPb

Tom is using Ron Miller’s cost algorithms plus some inputs from Les Waganer. The reactor is significantly larger than those being considered in ARIES-CS. The plasma major radius is 14.4 m, Baxis is 5.5 T, Pfusion is 4.5 GW, Pelectric is 1.94 GWe, <Pwall> is 2.1 MW/m2, and thermal conversion efficiency is 40%.