ARIES-CS Project Meeting Minutes
15-16 September 2005
Princeton Plasma Physics Laboratory, Princeton, NJ
Documented by L. Waganer
Ref: Agenda and Presentation Links: Project Meeting
Welcome - Rich Hawryluk welcomed the ARIES team and the Stellarator community that is reviewing the ARIES-Compact Stellarator (ARIES-CS) to the Princeton Plasma Physics Laboratory. There were several reviewers who participated from Europe via phone. Other community members have offered to review the presented material and forward their comments to the ARIES team. Our thanks to Jim Lyon and Mike Zarnstorff for arranging the meeting, soliciting the review members, providing the telecommunications facilities, and hosting the meeting. Our thanks to Pamela Hampton for an excellent job of facilitating the meeting, facility, and hotel arrangements.
Next Meeting and Conference Call - It was decided the next ARIES meeting would be held in San Diego on November 17 (Thursday) and the morning of November 18 (Friday). The next conference calls would be held on September 30 and October 18.
Status of ARIES Program - Farrokh Najmabadi noted that the ARIES-CS is 2/3rds completed. The detailed design assessment and optimization remains to be completed and the final report written by the end of the 2006 Fiscal Year. Al Opdenaker affirmed the urgency of completing and documenting the present study by the end of FY06. Farrokh noted that the final design might include several design options, if deemed appropriate.
ARIES-CS Town Meeting Review
Purpose of Town Meeting Review of ARIES-CS - Farrokh Najmabadi informed the review team that the intended purpose of the meeting is to review the assumptions, methodologies, and decisions made for the ARIES-CS involving the general areas of physics, engineering, and systems analysis. Based on the comments received, the team will make the appropriate corrections to proceed into the conceptual design phase.
Compact Stellarator Reactor Physics Design Basis
Attractive 2- and 3-Field Period Plasma and Coil Configurations – A Review of Progress and Status – Long Poe Ku explained the purpose of the study was to broaden and analyze the configuration design space to find improved reactor designs, including extensions of the NCSX approach and new Quasi-Axisymmetric (QA) approaches. Many configurations were examined to be able to implement these approaches into the systems code for parametric assessments for the optimal configuration. It is hoped these configurations will be competitive with other confinement approaches. For comparison, Long Poe illustrated a set of stellarator designs with a range of major and minor radii. The ARIES Compact Stellarator more closely approaches the size of tokamak designs rather than the prior conventional stellarator approaches.
Long-Poe described the approach and rationale for both the NCSX and the QA approaches. The QA approach included both MHH2 and SNS types. The present state of design development is preliminary with only the steady-state operating point being defined (startup and control not evaluated to date). Long Poe mentioned the codes used for the plasma and coil definition. Design criteria were set to be a noise content <2%, magnetic ripple <1%, and alpha energy loss <10%. The stability beta limits would be based on linear, ideal MHD theories. Ballooning modes should be stable to infinite-n modes and kink modes stable to n = 1 and 2. Shafranov shift would be set to ½ and flux loss due to isolated islands < 5%. The coil design criteria was established to assure sufficient coil separation, adequate radius of curvature, nominal complexity, and a minimum radius allowance for plasma, scrape off layer, internal components.
Long Poe then illustrated several plasma and coil configurations to describe the range investigated. Parametric charts of the maximum field versus the coil aspect ratio were plotted for several coil cross-sections.
Another configuration type was the N3ARE that had good QA, alpha confinement, and MHD stability characteristics. It also has lower effective helical ripples and better energetic particle confinement.
The KQ26Q configuration improved the robustness of flux surface integrity and good equilibrium flux surface quality, but has remnants of an m = 4 islands. This approach may have unstable free-boundary modes for beta greater than 4%.
MHH2 designs with aspect ratios of 2.65 have low field ripples and excellent confinement of alpha particles. A specific configuration, MHH2-K14 has an ultra-low aspect ratio, a similar quality of equilibrium similar to the NCSX. The KQ26Q configuration improved the robustness of the flux surface integrity. It has a good equilibrium flux surface quality, but the islands may cause some concern. Both of these designs were examined for field ripples and alpha confinement.
The SnS family of configurations had flat iota profiles and excellent flux surface integrity in QAS. The KJC167 configuration may be unstable to free-boundary modes with betas around 6%.
Summary – The NCSX class of configurations have been extended for better alpha confinement and surface integrity. More classes of configurations have been identified and developed for smaller aspect ratios, better QA and more robust surface quality. Design approaches have provided good separation of core components for maintenance actions. Many of these configurations have shown promise to be attractive compact stellarator power reactors.
Discussion of Physics Design Basis
Mike Zarnstorff commented about the coil designs. I believe he mentioned the NCSX coils were larger radially to accommodate neutral beams between the coils. If the reactor design does not require neutral beams, the coils could be more radially compact.
Basis and Applicability of Beta Limits to Compact Stellarators - Alan Turnbull asserted that the basis of MHD beta limits in stellarators is not as well understood by the fusion community as are the limits for tokamaks. The stellarator experiments appear to violate MHD stability limits, but both confinement concepts have the same underlying physics. Some insight may be gained by examining the tokamak limits. Equilibrium characterization is crucial to identify the problem. Stability can depend on the profile details.
The general consensus is that large stellarators (LHD and W7-AS) achieve beta values higher than predicted. The question arises then: should reactor studies ignore stability theory? A definite conclusion eludes the theorists until more data and analysis are conducted. Stellarator experiments are beginning to distinguish physical relevant instabilities with certain growth rates. Stability to these growth rates yield beta limits above a 5% beta in W7-AS. Low m/n values of 1/1 and 2/3 and 2/5 internal modes appear to establish the beta limit in LHD.
Several types of external modes need to be distinguished, such as global beta modes, current driven external kinks, current driven peeling modes, and ELMs. ELMs are significant because they are observed both in tokamaks and stellarators. Global edge stability depends strongly on edge conditions and rational edge values. However, the LHD experiment has a beta in excess of that predicted by low-n growth rates exceeding the critical threshold. The LHD is thought to have a resistive interchange mode as predicted by the observed edge MHD mode.
Equilibrium limits are still a prime candidate to establish an operational beta limit in stellarators. Equilibrium degradation may establish the W7-AS beta limit.
Summary - LHD and W7-AS have achieved beta values in excess of the Mercier interchange limit. LHD appears to have a beta limit established by m/n = 1/1 ideal limit. W7-AS seems to be limited by an approach to the equilibrium limit. Regardless, ideal stability plays a direct or indirect role in the degradation of the equilibrium associated with MHD stability limits and strongly growing ideal modes may provide direct limit.
Discussion of Beta Limits
Rich Hawryluk commented that the claim that "infernal" modes in Tokamaks do not generally result in disruptive beta limits was countered by cases in TFTR in which they did cause disruptions. ("Infernal" modes are a specific type of internal mode that pop up and disappear with seemingly little pattern as the minimum q-profile changes). However, Alan Turnbull pointed out that the names of modes were somewhat fluid and that since they extended across the entire plasma, he considered those modes in TFTR (and similar modes in DIII-D) under the category of global external kink modes. The "infernal" modes that are isolated to the low shear region are relatively benign. This actually provides an example of how the computed limits need to be interpreted.
Compact Stellarator Reactor Integrated Systems AssessmentSystems Code Results: Impact of Physics and Engineering Assumptions – Jim Lyon noted the approaches he used to define the reactor parameters in the systems code analysis: 1-D cost optimization, 1-D power balance for plasma parameters, and 0-D scoping study for device parameters. Parameter optimization integrates the plasma and coil geometries with the reactor constraints. Jim showed a table of configuration parameters for the six NCSX types and the two MHH2 configurations. These configurations illustrated that Bmax/Baxis depends on the coil cross-section. Laila El-Guebaly has stressed the need for a shield-only blanket region and the NCSX configurations do permit that type of blanket.
One of the coil design issues is the type of superconductor (NbTiTa versus Nb3Sn). Above 10 T and below 14 T, the NbTiTa superconductor has the advantage.
Jim described the 0-D scaling of the main reactor parameters. The neutron wall loading was maximized subject to the coil current density and the radial build constraints. The maximum wall loading is fixed at 5 MW/m2 peak. With a peaking factor of 1.52, this yields an average neutron wall load of 3.3 MW/m2. For a fusion power of 2 GW, the size of the reactor is determined based on a surface area of 480 m2. With these general constraints, the systems code optimizes the reactor parameters. Other variables are fixed also, including H-ISS95 <6, impurity levels, alpha particle loss at 20%, and profile shapes. Jim elaborated more on his treatment of plasma impurities.
Jim summarized a typical systems code output list of parameters. On n-T plots, he mapped out the operating point and the line of average beta = 6%. The operating points move higher temperatures with lower startup power as the ISS95 Multiplier H increases. Jim also presented a set of component cost numbers, but cautioned the cost estimates were in work and some costs were not validated yet. He provided some comparison of different coil configurations. The cost of electricity varied as 1/neutrom wall loading and a weak dependency with beta for two aspect ratios. The results seemed to be insensitive to profile assumptions.
The next step is to incorporate ν* as a target parameter, investigate neoclassical impurity profiles, analyze 1-D models of Te(r) and Ti(r), incorporate NbTiTa coil cost modeling, and analyze the field period maintenance approach.
Discussion of Systems Code Results and Assumptions
Dale Meade – Is the cost of the stellarator coils significantly higher than those for a tokamak? Answer: Presently the costs are nearly identical as the unit costs have the same cost and complexity factor. We are presently working to better define the coil design approach, materials, and costing methodology.
Dale Meade – The total weight of the nuclear island seems low. ITER was 23,000 tonnes. Answer: We agree it seems low and are looking into this. But the aspect ratio is higher and the size of the coils is smaller. There are no PF and EF coils and supporting structure as in tokamaks.
Compact Stellarator Reactor Engineering Assessment
Engineering Study of ARIES-CS Power Core and Maintenance – Rene Raffray explained how the team was using a the three phase assessment of the Compact Stellarator as a power plant. In 2003-2004, the study group explored the plasma and coil configuration and engineering options. In 2004-2005, the exploration extended to the configuration design space. In 2005-2006, the detailed systems design space will be explored and optimized. Several QAS configurations were explored, including NCSX-like three-field period configurations and 2-field configurations typical of MHH2. These two coil and plasma configurations were assessed with different maintenance approaches and blanket concepts in Phase I. In Phase II selected blanket concepts for both port- and sector-based maintenance schemes were assessed in greater detail.
Particular engineering emphasis was placed on the blankets and coil concepts. Five blanket concepts differentiated by structural, shielding, breeding, and cooling materials, as well as construction approaches. Key performance parameters for these five approaches were provided. Two approaches were selected for further study. The first was the dual coolant with self-cooled Pb-17Li zone and He-cooled RAFS was selected as the baseline blanket approach. The second was the self-cooled Pb-17Li blanket with SiC/SiC composite as the structural material. The dual-cooled blanket was redesigned from a prior design to be simpler and have a more effective coolant routing. Rene explained this blanket can employ the Brayton cycle to improve the thermo-conversion efficiency over a range of neutron wall loads and material properties.
The maintenance approaches include both field-period and port based maintenance. The port-based scheme includes a vacuum vessel internal to the coils and large ports to enable removal of large sections of the life-limited blanket components. There are 1 to 2 ports per each of the three field periods. Blanket modules are designed for port maintenance. Replacement design features for the field period replacement were discussed. The individual cryostats are enclosed in a common external vacuum vessel. Easy access and removal/replacement is designed into the coolant manifold piping. Extraction of the blanket components from the stellarator field period fixed shield was verified with CAD models to ensure no interferences.
The coils and coil structures are designed to accommodate both types of maintenance schemes. It was determined it is more efficient to embed the coils into a continuous stellarator torus tube. The thickness will be varied according to the local stress levels. The coils have been designed with FEM modeling as will the coil structure in the future. Magnetic flux densities were determined to determine the maximum flux density.
The divertor is a critical part of any stellarator. The initial physics study established the divertor parameters to locate the divertor plate and determine the heat load peaking factor. These parameters are considered to be preliminary and will continue to be refined. A divertor component has been designed to meet the design requirements of 10 MW/m2. The configuration is in the shape of a T-tube divertor. The divertor will be manifolded and integrated into the blanket coolant system.
Discussion of Engineering Assessment of Power Core and MaintenanceMike Zarnstorff: How do you support the gravity loads through the vacuum vessel and cryostat and is the bucking cylinder at cryogenic temperature? Answer: We had designs for the gravity supports to be long and have a low heat conduction to the base. We will consider making the bucking cylinder warm structure.
Compact Stellarator Assessment
Discussion of ARIES Compact Stellarator
Hutch Neilson: This presentation was very useful. We learned a lot about your Compact Stellarator conceptual design approach. Most of the concerns were addressed during the presentations.++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
ARIES-CS Project Meeting
Compact Stellarator Reactor Physics Basis
Next Steps on Plasma and Coil Configurations – Long Poe Ku noted that his presentation at the Town Meeting covered the current status and the near term work areas. He did not prepare a separate presentation for this project meeting session. He reviewed his efforts needed to support the definition of the divertor and the coils supporting the Engineering efforts.
Discussion Summary: Basis and Applicability of Beta Limits to Compact Stellarators – Alan Turnbull summarized the outstanding questions about beta limits. Are nested surfaces a valid approximation for stability calculations? Non-linear consequences are crucial for interpreting stability calculations. How rigorous is the link between approach to stability limits and equilibrium degradation? Alan says he is not completely sure all of these have been resolved.
Alan then presented his planned procedure to help mature the CS design: ignore local stability criteria, check linear global stability, monitor linear stability predictions against non-linear prediction, and check flux surface quality. Further input from the stellarator community and experimental evidence is needed for higher confidence. Additional correlation between tokamaks and stellarators to more fully understand the beta limit theory.
An Update on Divertor Design and Heat Load Analysis – TK Mau reviewed his divertor design criteria. It consists of a number of divertor plates between the LCMS and the first wall to intercept and remove the heat flux and the peaking factor. He also listed the computer codes used to calculate heat flux and geometry.
The Poincaré plot of the field lines outside the LCMS provides data for the placement of the divertor plates and the peaking factor. At present, the peaking factor is too high at around 20:1. For the latest design, approximately 30,000 lines were followed to intercept locations. TK showed some typical results of the divertor design. The divertor plates are roughly conformal to the LCMS at one extremity of the plasma surface (small radius of curvature). The plates have some asymmetric tilt to the LCMS. TK showed a distribution of heat flux to help identify the extent of the divertor plate and the local heat flux peaking. The field line angle of incidence is relatively small – the average is 3.3º with the maximum being 9.9º. The field line length is crucial in determining the scrape off layer parameters. Some conduction power does reach the first wall. TK provided some data on the projected peak loads on the divertor of 22 MW/m2. Some work remains on the alpha particle heat flux determination.
Farrokh Najmabadi was concerned that the divertor design has not converged on an approach that is consistent with the desired criteria. The design inputs are not consistent throughout the team. We need to establish a single consistent reference design. Long Poe should provide a baseline coil set that can be used by UCSD to establish heat flux and divertor design parameters. This would also benefit the Systems Code definition. This should be in place by the November meeting so we can achieve some design definition progress. See Action Item List at the end of the minutes for more detailed action definition.
Compact Stellarator Reactor Integrated Systems Assessment
Recent Results and Next Steps for the Systems Code – Jim Lyon noted that the presentation he gave at the Town Meeting contained the most current results. He noted the suggestions for the action items will be helpful to converge on a recommended design.
Compact Stellarator Reactor Engineering Assessment
Stellarator Magnet Options – Leslie Bromberg highlighted the electrical and winding issues. The ARIES-AT high temperature magnets do not apply well to the ARIES-CS application because of the quench properties. A good mechanical and thermal contact between the HTS and the structure makes quench energetically impossible with high temperature and energy margins. Leslie suggested the use either a low temperature superconductor like NbTi and wind/react like Nb3Sn or MgB2. Leslie showed the SC winding pack current density for both Nb3Sn and NbTiTa. He further noted the remaining issues with these superconductors, such as the winding process in the CS configuration, and material definition of the inorganic insulator. Quench protection is still an issue. It was also brought up that the coil structure definition and cost needs to be determined. The bucking cylinder needs further definition as a warm structure.
Status of the Coil Structure Design and Magnetic-Structural Analysis – Xueren Wang highlighted the main challenges for designing coil structures to satisfy both the modular and field-period maintenance approaches: coil support, force reaction, and transition from cold to warm structures. He presented the arrangement of the coils for an entire field period as one integral supporting tube that houses the coils.
Electromagnetic and structural analyses have been conducted on the provided ARIES-CS coils using FEM models in the EM analysis. Xueren showed the results of the EM analyses for the current density distribution and the magnetic flux densities with the resultant forces.
Structural analysis on the bucking cylinder has been accomplished – Xueren provided the structural analysis results and material properties used.
Status of the ARIES-CS Power Core Configuration and Maintenance – Xueren Wang presented the approach for the replacement of modular maintenance of the dual-coolant liquid lead blanket modules. He explained the design approach for the access pipes that connect the manifolds to the blanket modules. One of the key issues is how to protect the welds between the access pipes and the manifold. There will be a lot of neutron streaming through the access pipes that degrade the ability to reweld the access pipes. Laila El-Guebaly provided the radial builds that describe the configuration.
Xueren described the replacement units for the field period maintenance with the re-usable HT shield and manifold. Further descriptions were provided for the access pipe configuration and weld joints. Expanded 3-D drawings were shown. Xueren explained that toroidal movement of the replacement units had some conflicts. These might be resolved by reducing the radial depth of the breeding zone or local replacement of the breeding zones by shield-only zones. He showed cuts through the power core to illustrate the clearances needed.
Status of ARIES-CS Power Core and Divertor Design and Structural Analysis – Rene Raffray focused on the engineering effort results from Phase II and the action items required to resolve outstanding issues. He provided designs obtained for the divertor elements and how it integrates with the blanket module and its coolant systems. He summarized the coil structure design and analysis that has been performed. Other ancillary is needed but not investigated to any degree at this point. He also outlined the dual coolant modular design and its performance characteristics.
Rene summarized the options for selection for the blanket and maintenance schemes. He did outline a suggested procedure to select the final design choices.
2-D Analysis of Neutron Streaming Through Helium Access Tubes – Laila El-Guebaly highlighted issues and concerns for the re-weldability of the manifolds and the neutron streaming through the helium access tubes. As reference, she showed the reference radial builds and a modified radial build through the helium access tubes. She noted the differences between field period and port maintenance approaches. She showed Siegfried Malang’s innovative manifold to improve the access tube weld properties. She provided details of the neutronic effects at specific locations and suggested certain improvements. She summarized her 2-D results of these weldment zones. She also provided the damage profile at the vacuum vessel as well as the fluence and heating at the magnet on the tube centerline. Neutron streaming through the 206 He access tubes results in hot spots where the damage exceeds the limit by roughly an order of magnitude. Local shields are needed to protect the vacuum vessel and magnet.
It became apparent during the talks that the team had not used a consistent physics and engineering baseline. The following action items were established, primarily to make sure the analyses and design development would represent a consistent approach. This is especially necessary as there is a set of technical papers to be prepared and delivered shortly. The action items are: