ARIES Program
Public Information Site

ARIES Documents -- Meetings Archive

ARIES Conference Call, 25 February 2003

Documented by L. Waganer

(ANL) -
(Boeing) Waganer
(DOE) -
(FPA) -
(GA) Turnbull
(GT) Abdel-Khalik
(INEEL) Merrill (Brad)
(LANL) -
(LBNL) -
(LLNL) -
(MIT) Bromberg
(MRC) -
(NRL) -
(NYU) Garabedian
(ORNL) Lyon, Spong
(PPPL) Ku, Mynick, Neilson, Schmidt
(RPI) Steiner
(SNL) -
(UCSD) Mau, Raffray, Tillack, Sze, Wang
(UW) El-Guebaly

Ref: Agenda and Links


Les Waganer and René Raffray reviewed the general plans for May ARIES town meeting and project meeting to be held in Livermore near LLNL on May 5-8, 2003. The ARIES Town Meeting will feature an in-depth discussion of IFE Liquid Wall Chamber Dynamics.

This conference call is intended to convey more detailed technical information than a regular conference call. Several electronic presentations were uploaded onto the ARIES web site per the above link. As the presenters discussed each viewgraph, the call participants could view the viewgraphs via the web link. Presentations may be downloaded for reference.

Status of Compact Stellarator (CS) Technical Areas

Hutch Neilson and John Schmidt noted that Phil Heitzenroeder, Wayne Reiersen, and Long-Poe Ku provided the ARIES team an initial set of stellarator engineering and physics data representative of a reactor-grade plasma and power core system parameters. Subsequently, this data have been placed on the ARIES web for Team access.

Recent Progress in Configuration Development for Compact Stellarator Reactors - Long-Poe Ku summarized the progress obtained since January 2003 that will be discussed in the presentation. The reactor configuration is based on the NCSX stellarator design. The modular coils, the M50 configuration, and the related reference plasma were scaled to a size expected to produce a fusion power consistent with a net power output of 1000 MWe. This provided an initial starting point for the team to begin their engineering studies. One caveat is that the M50 coil design is based on copper conductors cooled with LN2 as opposed to superconducting or high temperature superconducting coils. The design parameter are R = 8.25 m, A = 4.5, B = 5.3 T, Beta = 4.1%, and Ip = 3.5 MA. This configuration would yield a mid-coil to plasma distance (D) of 1.2 m and a neutron wall loading, NWL, = 2 MW/m2, not the 1.4 MW/m2 reported previously. It was emphasized that these values are just starting points.

A new tool provides a two-D map of the coil-to-plasma spacing. Two islands of D-values with minimum values of 1.2 m were seen per field period. Previous recommendations from the Engineering Group suggested a minimum value of 1.45 m would be allowable. Therefore, Long-Poe determined there was around 8% of the FW area at a distance of less than 1.45 m. Laila El-Guebaly separately determined this value to be in the range of 10%.

Long-Poe also illustrated results from a code to calculate the magnetic field intensity and current density in the coils. These are average values assuming the coils to be normal coils.

The search for an attractive reactor plasma configuration (high beta, high field, good alpha particle confinement) was widened to include a range of aspect ratios, Iota, and field periods about the nominal A = 1.5 and three field periods. To help understand the stellarator physics involved, Long-Poe examined pressure profiles and bootstrap current density regimes. Certain pressure profiles yield plasmas with reduced kink stabilities. One question is, what beta level should be a goal? Higher beta values lead to higher wall loadings, which is desirable for a reactor-grade plasma up to some threshold value. A neutron wall load of 4 MW/m2 might be a reasonable goal. Another possibility is to increase the reactor size from the nominal 1 GWe up to 2 GWe.

A question was raised about the possibility of a liquid metal surface for the first wall. René Raffray said the currently envisioned solid first wall concepts should be able to handle the peak neutron and heat flux wall loading being considered. It might be reasonable to consider liquid surfaces on the divertor regions but the liquid should be compatible with the blanket materials.

Stellarator Reactor Optimization Code - Jim Lyon and Don Spong related that ORNL has limited funding for this effort and want to make sure it is properly allocated to provide the maximum benefit to the CS reactor project.

Jim first discussed his 0-D spreadsheet stellarator reactor optimization code that has fixed plasma and coil geometries. By varying key input parameters, the code minimizes the major radius for a target fusion power level. It is useful for size scaling for fixed plasma and coil geometries and comparison of reactor configurations.

More time was spent on the 1-D optimization approach that uses several optimization codes linked together. Jim described how the codes operate for a complete optimization that is computer time intensive. A more time efficient approach is to optimize the plasma and coil geometry before entering the more intensive MHHOPT code. Step-wise optimization might be a faster way to converge to the desired optimal reactor configuration. Jim then detailed the input and output parameters for each of the optimization code elements.

Jim discussed the alpha confinement issues and their impact of key reactor parameters, such as power balance and localized heating loads. Jim presented a view of the plasma with a plot showing where the fast ions exited the last close surface. The alphas were concentrated on and along the outer, sharper edges of the plasma. Jim discussed several methods to optimize (reduce and/or relocate) these alpha losses.

Jim and Don are planning to upgrade the current math libraries and operating system. Then plasma and modular coil geometries and upgraded costing algorithms can be incorporated in the codes. Alpha particle loss regions and divertor areas will be investigated. The COILOPT/STELLOPT will be incorporated in the optimization. The group asked if Jim could run the Long-Poe parameters and coils presented earlier - Yes, he could. Could he complete it by the next meeting - Yes, he thought he could, as well as converting the coils to be superconducting. He would post a set of references for his codes on the ARIES web.

ECRF Heating - TK Mau looked at the new Long Poe Ku data relating to ECRH system and concluded they were generally similar to those obtained at the last meeting and sufficiently correct to continue using them. Earlier, he talked to the Wendelstein group about ECRH and they are very interested and would like to collaborate with him.

TK is investigating ways to improve the simple calculation process he is currently using. He hopes to have preliminary results and ECRH system implications by the next meeting.

Magnet Engineering - Leslie Bromberg thought the high temperature superconductor would be feasible in this application. If not, Nb3Sn coils could be employed. He would like to move the vacuum vessel outside the winding pack and replace the VV with a thinner shield. Laila did not think this would be possible, or beneficial, as the radial build is already well optimized. They agreed to discuss this change off-line. Leslie proposed a 20 cm thick low-temperature S/C magnet and did think there would be no problem achieving a mid-coil to plasma distance (D) of 1.25 m. This was a very important finding, as this may eliminate the need for a shield-only region and also simplify the blanket with only one type of blanket.

Radial Build Updates for FLiBe/FS System -Laila El-Guebaly reported on the design changes on the radial build accomplished and the implication of these changes. Previously the old average neutron wall loading was around 1.4 MW/m2 with peaks around 2 MW/m2. The new Long-Poe Ku design is around 2 and 3 MW/m2 for the average and peak, respectively. To protect the S/C magnet, the vacuum vessel should be 25 cm thick and cooled with borated water. For the 20-cm-thick magnet, the closest approach is reduced from around 1.25 to 1.05 m in two areas of roughly 8-10% of the total surface area. If this area has to be the Shield-Only blanket type for minimal thickness regions, this would require adjustment of the tritium breeding, as well as using more costly WC shielding materials. There is no impact on the power balance, as the heat will be recovered from all components.

Laila provided some design parameters for the coils and the composition of the power core components for both the regular blanket and the shield-only blanket regions. She had some plasma and first wall cross-sections that illustrated the location and field-period fraction containing the minimum coil/plasma distance and the highest neutron wall loading. These figures highlighted the magnitude of the minimum distance problem. Laila presented several prior blanket radial thicknesses to illustrate that this will be an aggressive design. However, the 20 cm saving in radial build could have significant impact on the overall machine. To help understand the magnitude of the tradeoff, Jim Lyon agreed to run 3 cases for Dm using the systems code: 1.05, 1.25, and 1.45 m.