ARIES Program
Public Information Site


ARIES-CS Project Meeting Minutes

14-15 June 2005

University of Wisconsin, Madison

Documented by L. Waganer

Organization ARIES Compact Stellarator
Boeing Waganer
FXK Ihli
General Atomics Turnbull
Georgia Tech Abdel-Khalik, Yoda
INL Merrill
RPI McGuinness, Steiner
UCSD Mau, Najmabadi, Raffray, Wang (Xueren)
UW-Mad Blanchard, El-Guebaly, Henderson, Kulcinski, Martin, Santarius, Sawan, Sviatoslavsky, Tautges, Wang (M.)

Ref: Agenda and Presentation Links: Project Meeting


Welcome - Laila El-Guebaly informed the team about the facilities and eating arrangements at the University of Wisconsin, Madison. Jerry Kulcinski gave the team a warm welcome to the University, although he did not offer to provide any of the great fish he caught during his fishing trip. Our thanks to Laila and Jerry for hosting the meeting.

Status of ARIES Program - Farrokh Najmabadi understands the FY06 Presidential budget is roughly the same as the present year, implying no or minimal cuts to the Systems Studies budget. However, the impact of ITER commencing will have some effect on all budgets.

The next ARIES meeting is to be held at PPPL. Farrokh had tentatively selected a date during the week of 19-23 September for 2 1/2 days, but the specific dates are TBD.

Farrokh said the ARIES team must be technically ready (final design approach and supporting analyses) as this review meeting will be an in-depth assessment of the ARIES Compact Stellarator approach. A wide spectrum of stellarator advocates and knowledgeable experts will be invited to attend in-person and electronically via a web-based meeting. In addition to the baseline approach, the design space should be determined sufficiently to enable the selection of the optimal design point. This implies that the project must have the plasma/coil configurations, design approaches, and system study parameter assessments to choose the design parameters.


Summary of IEA Socio-Economic Workshop - Les Waganer recently attended the IEA Socio-Economic Workshop, along with Jerry Kulcinski, John Schmidt, and Farrokh Najmabadi. The purpose of the workshop is to reach a consensus on the socio-economic aspects of fusion as a commercial power plant over the next 50 years. The demand for energy will continue to escalate for all developed and developing countries, especially China and India. The greenhouse effect is expected to be addressed in the upcoming G8 Conference with increasing carbon dioxide sequestration and/or taxation. The use of fossil fuels will increase until resources and CO2 storage becomes too great a burden. Use of renewable energy resources will also increase, but the resources, geographic limitations, and costs (with storage) will be limiting. Fission will see an increased use with rebirth in UK and USA and strong use in the remainder of EU, Japan, and China. Fusion will not be in significant use until beyond 2060 and only then if CO2 limitations are invoked. Some EU representatives feel the external costs should be used to evaluate all energy sources, thus improving the competitiveness of fusion. On the other hand, the US believes external costs should not be explicitly included and fusion should be demonstrated to have COE values lower than competitive sources to offset the specter of risky technology.

Summary of ISFNT Conference - Laila El-Guebaly summarized the points of interest from the ISFNT Conference for the ARIES study. The EU presented their design concept for the DEMO and power plant studies. For the helium- cooled LiPb design, the helium pumping power was stated to be 400 MWe, which might be too high. However, a majority of the expended pumping power could be recovered for energy conversion. Voss (UKAEA) suggested using Wcoated SiC pebbles to cool the divertor. Europeans are also developing advanced ODS EUROFER to operate at 650C. The Japanese are assessing the supercritical CO2 as a coolant aiming for high thermal conversion efficiency.

Compact Stellarator Reactor Integrated Systems Assessment

Status of Systems Code Studies - Jim Lyon discussed his code changes since the last meeting. Five new coil configurations were added (two NCSX-like and three MHH2 versions) along with a revised radial build model. Jim noted that all volume calculations need to be correlated with the CAD model and the distance from the VV outer surface to the coils needs to be verified for all blanket regions. He is still concerned about the definition of the coil structure and the cost algorithms for costing the coil pack and structure. The power flow diagram needs to be validated. The code did converge on a constrained solution.

Jim presented a revised set of plant performance and design values and costs at varying levels of detail. These costs and component masses were compared with prior ARIES designs. There remain a few costs that look suspicious. Les Waganer should reverify the unit costs and convert them to the year dollars Jim is using (2004$). Laila will check and verify the cost accounts 21-26.

Jim summarized the remaining code improvements, such as adding the divertor definition, improved coil support structure, VV costing, shielding thickness, improved port maintenance modeling, field period maintenance modeling, and external VV approach.

Compact Stellarator Reactor Physics Basis

Assessment of New 2-FP and 3-FP Plasma and Coil Configurations Long Poe Ku presented analyses for two new, A=2.65, two-field period configurations, MHH2-1104 and K-14, that were discussed at the last meeting. These have low ripple characteristics and excellent α -particle confinement, but different total rotational transforms. Long Poe has applied more rigorous design optimizations that led to new coil designs. Thus, it has been demonstrated α -loss design can be achieved with discrete MHH2 coils at finite β that would lead to an attractive compact stellarator. Long Poe then presented his analysis results of these two concepts.

Long Poe also analyzed the three-field coil configuration called KJC17 for equilibrium, field ripple, α-loss, and flux surfaces. He then compared the characteristics of this design to the prior configuration, NCSX-M50.

In the future, he will re-examine the NCSX configurations to further improve the energetic particles and further optimize the MHH2 configurations.

Update on Beta Limits Alan Turnbull continued his exploration of the β-limits as applied to compact stellarators, specifically on ballooning modes, global internal and external modes, resistive interchange modes, and equilibrium limits to β. It is generally conceded all large CS experiments routinely exceed local ideal ballooning and Mercier β-limits, but local modes can also appear where Mercier is predicted to be stable. Strict tokamak limits are generally not applicable. But there are also precedents for ignoring some of the same ideal stability thresholds in tokamaks. W7X is providing some meaning to ideal global internal modes. The β-limits for low n-modes (m=14) is about 6%, whereas β-limits for higher m-values are 5.25%.

In LHD, the low m/n modes are excited in the edge region and are stabilized as β increases. Strong toroidal shear may stabilize the 1/1 mode in NSTX.

The ideal MHD stability can affect the limits in two ways: 1) degradation of equilibrium and 2) growing ideal modes to provide direct limit.

MCNPX/CAD Activities and Preliminary 3-D Results Mengkuo Wang explained the improving capability to accomplish 3-D neutronics on very complicated geometries, as driven both by ITER and ARIES-CS. Researchers at Sandia National Laboratories and the University of Wisconsin-Madison have overcome a major limitation in Monte Carlo-based radiation transport, by integrating the Monte Carlo radiation transport code MCNPX directly with a CAD geometry engine (CGM). There are differing approaches between translation methods (inexact solutions) versus direct methods (exact solution from the CAD geometry). Mengkuo noted that he has successfully constructed the exact CS power core geometries out to the coolant manifold region. The run time has been reduced from days to one hour. He explained his process of surface generation and incorporation of the radial builds into the model. The 3-D neutron wall loading results had close agreement with the prior 2-D results. His future plan includes higher fidelity exchange of CAD data and incorporation of the divertor and penetrations into the model to estimate the overall TBR and the breeding in each distinct blanket zone.

Divertor Heat Load Analysis - TK Mau showed his diagram modeling of the power core heat flow to obtain a better understanding of the criterion for the divertor. He used the W7-X magnetic field model to understand the a-particle gyro orbits using GYRO. He reported some progress from Hayden McGuinness on locating the divertor plates. TK is modeling the heat flux from the non-thermal alpha losses. He also tested GYRO on the NCSX CS magnetic geometry.

In the future, an algorithm will be developed to use the GYRO results from drift orbit calculations to start tracking when particles reach the LCMS. Features in GORDON will be interfaced with GYRO to determine exit points from LCMS to the FW.

Possible Configurations for Divertor Hayden McGuinness outlined his methodology on determining the heat loads to the divertor region so that the region can be quantified. He described three cases that he analyzed in some detail:

  • Wall conformal to LCMS
  • True conformal wall
  • Wall defined by radial field lines

He felt the ergodic design was robust. Design details are important in small dimensions. Plates are not necessarily conformal. Tilting with the field lines is beneficial. The high line interception is achievable.

Compact Stellarator Reactor Engineering Assessment

Status of Divertor Design and Structural Analysis Thomas Ihli presented the design details of his evolving helium-cooled T-tube divertor concept. The coolant is introduced into the divertor through a coaxial connection into the internal part of the main coaxial tube. It is then forced from a slot in the inner tube onto the surface of the outer tube to which is attached the divertor W-surface. He thermally modeled the components of the divertor approach and determined the stress intensities and deflections throughout the divertor. A very high effective coefficient of heat transfer can be obtained with this approach. The nominal heat load is 10 MW/m2, although that number can be altered. The transition from the steel anchor to the tungsten is thought to employ layers of varying FS and W compositions.

Details of the divertor design approach were shown. Laila is concerned about neutron streaming along the helium coolant channels through the thickness of the blanket and shield. These channels will have to be tailored to minimize this undesirable effect.

Thermal Analysis of Helium-Cooled Divertor Said Abdel-Khalik stated the objectives to evaluate the thermal analysis of the T-tube, helium-cooled divertor at nominal conditions and conduct parametric studies of the geometry and operational conditions to assess the sensitivity of the thermal performance. Both 2-D and 3-D analyses have been done for nominal operating conditions and dependence on temperature distributions and pressure drops. The 2-D geometry has been simulated with a high fidelity gridding system. The 3-D geometry of the coaxial tube is modeled to be able to accurately solve the conservation equations. Pressure and velocity results for both the 2-D and 3-D models were provided. Parametric variations for the 2-D analysis were shown for turbulence model, mass flow rate, and inlet pressure as well as slot width and jet-wall spacing. The 3-D thermal model showed the highest temperatures were on the tungsten surface at the edges of the limiter. Parametric results for the 3-D model were also provided. It was concluded that the maximum calculated temperatures at nominal conditions and design parameters were as predicted. The sensitivity studies indicated the design approach is robust.

Status of Engineering Assessment of Power Core and Maintenance Rene Raffray said the major engineering effort was on the divertor definition. Work also progressed on the dual coolant blanket design with LiPb self-cooled blanket and He-cooled RAFS structure. The structure for the coil remains an unresolved engineering issue. The two maintenance concepts are still viable and are being developed. These and other concepts are being tracked along with the action items from the last meeting.

Rene reported good progress on the definition of the divertor, both in requirements definition and design approach. The coil structural analysis needs a higher level of definition to effectively integrate it with the other systems and support the systems code. The tritium extraction and recovery system needs definition. The heat transfer and transport system including the intermediate heat exchanger needs to be defined to be compatible with the LiPb and helium heat transport fluids.

Rene updated the dual coolant module design status. The thermal analysis has been completed for the heat transport loop including the Brayton power conversion efficiency. The plenum and plumbing has been redesigned to improve the efficiency. Presently, the RAFS wall temperature limit is 550C and the compatibility limit is ~500C to achieve an efficiency of 42-43% efficiency. Some parameters could be adjusted to achieve even higher temperatures. Production of hydrogen is a good alternative use of the thermal fusion power

Nuclear Issues for LiPb/FS/He System with Internal VV Laila El-Guebaly summarized the current radial build and compositions for the nominal blanket, the WC-shield only, and the transition section. She addressed the neutron streaming issues arising from the He/LiPb access tubes. The 18-cm thick shield provides sufficient shielding for the VV and magnet coils, but it does not allow reweldability for the manifolds. There are 206 access tubes for each coolant with approximately 4 He and 4 LiPb access tubes per meter. Damage behind the He access tubes could be excessive and 2-D analyses are needed to confirm the 1-D analysis results, which predicted the He production at the manifolds and VV and high fast neutron fluence at the magnets. The thermal power split between LiPb and He was computed to be 58:42. Laila also predicted the LiPb decay heat and the volume of the radioactive waste.

Shielding Options for FP Maintenance Approach with External VV Laila El-Guebaly outlined the shielding considerations for the LiPb/FS/He blanket option. The low temperature (LT) shield could be cooled with either He or H20. However, the water presents a potential safety concern for a water leak in proximity of the high temperature LiPb coolant. The best candidate filler is borated-FS. Laila then presented an optimization on the LT shield with both coolants that compared thicknesses and costs. She then recommended the FS structure with borated-FS filler and water coolant. With this choice, she displayed her recommended radial build for blanket and shields. A robust LT shield design with 30% FS structure should be developed to warrant a leak-free system. Les Waganer needs to reverify all the unit costs for the power core (excluding magnets) and convert them to 2004 $ for use in the Systems Code. Specific materials to be investigated are TiH1.7, B4C, and fabricability of SiC.

Status of Modular and Field Period Replacement Maintenance Xueren Wang has extended his 3-D CAD modeling to include the bucking cylinder, coils and coil structure, internal vacuum vessel, manifolds, hot shield, FW and blanket, and the confinement building. He continues to compute interferences and clearances and volumes for the systems code.

With the CAD drawings of the bucking cylinder, he illustrated how the cylinder would be assembled. Around the bucking cylinder, the three coil supporting structure elements would be added. Then the vacuum vessel elements would be added. After the VV is completed, the internal blanket, shield, manifold, and divertor elements could be added. The space restrictions around the tungsten shield-only sections are very limited for the manifolds and more design effort will be required to define these elements. Xueren provided the volumes of the key power core components.

Tritium Permeation and Extraction Issues for the ARIES-CS - Brad Merrill reviewed the tritium inventory and permeation safety limits for ARIES-CS. Site boundary limitations are typical of those for other ARIES fusion commercial power plant studies, namely a dose limit of 10 mSv. Other limits were discussed and compared to those expected in ITER. Brad discussed the factors that influence the tritium inventory and permeation rates, such as use of reduced activation ferritic steel, choices in heat exchanger materials, and gettering approaches and materials. The tritium inventory in the heat exchanger is dependent on pressure of the LiPb and the HX tube materials. Extraction flow rates were compared in the different blanket design options. Brad did not think the alkaline metal HX was a good extraction method. Extraction tubes also were not a good choice. Brad then discussed the recommended vacuum permeation method.

Brad discussed some tritium permeation safety issues relating the ARIES CS design approach, including sources of airborne and water-borne operational releases, permeation through the LiPb piping, and Brayton cycle helium leak rate issues. Many more detailed safety analyses will be needed after the relevant design features have been finalized.

Updated LOCA/LOFA Analyses for Blanket and Shield Only Regions Carl Martin has been updating the LOCA and LOFA analyses to reflect the latest radial build configurations. From the radial builds, he constructed finite element models for the analyses. Carl presented the thermal LOFA results for LiPb and water and LOCA results for He coolant. The maximum temperature excursion is 28C lower for the new configuration. The new LiPb decay heat data are lower than previous data scaled from ARIES-AT. The thermal response for LOCA in the blanket and shield plus a LOFA in the vacuum vessel results in a first wall temperature that is 5C higher than the previous estimate.

Carl also updated the WC shield-only zone FEM to include radiation from the first wall wall across the plasma to the relatively colder blanket region. Then in the LOCA, the result from the thermal analysis is a maximum temperature of ~1000C that is 368C lower than the previously reported, but still exceeds the maximum reuse temperature for FS (740C). The effect of radiating/conducting the decay heat toroidally to the colder surrounding blankets will be assessed.