ARIES-CS Project Meeting Minutes
27-28 April 2006
University of Wisconsin - Madison
Documented by L. Waganer
Ref: Agenda and Presentation Links: Project Meeting Agenda
Welcome - Dean Jerry Kulcinski welcomed the ARIES Team to the University of Wisconsin - Madison. He informed the team of the University facilities and the strong fusion program at Wisconsin. He discussed the advantages of returning to the Moon to harvest the Helium-3 for use in attractive fusion power plants.
Status of ARIES Program - Farrokh Najmabadi noted the need to conclude the design and analysis of the Compact Stellarator to meet the scheduled project end date.
The date and location for the next ARIES meeting was discussed and finalized as the June 14 starting at Noon and concluding on Thursday June 15. It was suggested ARIES team prepare several papers relating to the current study on the Compact Stellarator. The next conference call will be on Tuesday May 16. Les Waganer has informed the team of the conference call date and the next meeting date and location.
Note: Action items follow the main text.
Compact Stellarator Reactor Integrated Systems Assessment
Assessment of Power Core Parameters and Related Costs (Part III)- Laila El-Guebaly continued the assessment of the key ARIES-CS power core parameters and costs by comparing the cost of the reactor building of the -RS, -AT, and -CS conceptual designs. The new CS reactor building volume is rather low, but the definition should be considered as incomplete, so the volume and cost will likely rise (Action Wang and Waganer). Laila also provided data on the Primary Structure. It might be helpful to provide some definition of this primary structure rather than a fraction of power core mass (Action Wang). Laila questioned why the Vacuum Pumping System is so expensive and asked if it could be assessed (Action Najmabadi). Laila provided the scaling law for the shielding versus the neutron wall loading. The cryogenic heat load to the magnet was provided (~5 MW). She also showed the sensitivity of the COE to both availability and LSA. The ARIES team tentatively fixed the availability at 85 percent pending an availability analysis (Action Waganer). LSA will probably remain at 2 for the baseline FWBS material selection. The 7.5 m CS configuration produces much less radwaste than the 8.25 m machine, but still the 7.5 or 7.75-m CS machine produces more radwaste than other comparably sized machines.
Systems Code Status - Jim Lyon updated the team on his progress since the last meeting. He summarized his completed action items: used 50 MW Miscellaneous Power + 5 MW for Cyro power; adopted the 5 cm thick cryostat; used 76% availability but will increase to 85% pending Les' update and other key parameter inputs. He has changed several parts of the code to reflect the most current FWBS configuration, optimized the Fe% for radiation power requirements, and added cryostat definition.
Jim showed a graph depicting the parameters affecting the major radius, R. There was a desire to lower the value of R to 7.5 m, but this seemed to be too close to the limit on many parameters and it was felt that it would be better to adopt a more conservative R of 7.75 m (with respect to TBR, Li-6 enrichment, and geometry constraints). Laila pointed out the mismatch between the delta-min from the 7.75 m case (1.21 m) and the recommended value by Long Poe (1.3 m). Jim will check and fix the code error.
The power flow diagram was shown and Laila noticed that the alpha loss was not added into the divertor heat load - Jim will check if the code or the chart was in error.
There was discussion on the radiated power fraction and where the power ended up. It would be desirable to calculate a neutron source profile for P. Wilson for his NWL calculations and for TK Mau for his divertor definition. (Action Lyon)
The CS plasma probably will need some VF or control coils for startup and Jim needs the definition to add to the code. (Action Ku)
Replacement component definition and cost needs to be defined by Jim Lyon and Laila El-Guebaly (Action).
Compact Stellarator Physics Basis
New Developments in Plasma and Coil Configurations - Long Poe Ku recapped the January 06 presentations and discussions that emphasized the N3ARE configuration is a robust and valid solution. Particle confinement is highly dependent on device size. Long Poe continued to show the key physics attributes of the N3ARE configuration (low ripple, stable, and low alpha losses.)
Since January, Long Poe has investigated other similar or derivative configurations that promise improved physics (lower ripples and alpha losses). Rather insignificant shape changes suggest performance improvements. The aspect ratio can be lowered from 4.5 to around 4 in the ARE case that will improve the alpha confinement. This could be investigated in the systems code to see if there are any systematic improvements. The second area was to improve alpha confinement by increasing the core transform, e.g., a modified configuration called N3ASDE. An energy loss of less than 5% may be achievable with this configuration. This configuration will continue to be analyzed although there are no coil configurations at this time.
Update on Beta Limits for ARIES-CS - Alan Turnbull reviewed the problem with the analytical wall construction that caused Terpsichore to previously fail, i.e., sharp corners. He devised a new mathematical approach to construct a conformal wall that is compatible with the Terpsichore code. This approach yielded reasonable wall representation in all toroidal planes. Tony Cooper is supplying the necessary code changes to incorporate the new wall representation approach. This will enable investigation of the baseline configuration.
Divertor Heat Load Analysis and Derived Geometry - TK Mau mentioned the two available baselines he is modeling, the ARE at R = 7 m and the KZD at R = 8.25 m. He should begin now to model the adopted ARE baseline of R = 7.75 m (action). TK described the analytical model he is using for the divertor plate geometry with examples of the toroidal extent of the divertor at three poloidal locations for both conformal and flat plate divertor concepts.
TK noted that the computational time for his analysis runs exceeds the allotted computer time regardless of the number of points analyzed. Something must be wrong with the code that will not allow convergence.
Need to find a divertor solution that defines physical extent of plates (poloidal and toroidal), distance to LCMS, and both average and peak heat loads. (Action Mau).
Compact Stellarator Reactor Engineering Assessment
Near Final Radial Build and Nuclear Parameters - Laila El-Guebaly reviewed her previous action items to highlight those completed and remaining items. She stressed the 7 m machine is not big enough to provide adequate tritium breeding; hence, we should adopt either a 7.5 or 7.75 m device. She showed the new plasma-coil separation contours with the minimum separation distances noted. Her estimate of the blanket coverage was confirmed by Xueren's CAD analysis. For the 7-m case, she provided the estimate of uniform and non-uniform blanket coverage (65% vs. 35%). For the 7.5 m case, the values become 72% and 28%. The 8-m case rises to 80/20.
Laila presented her current, near-final radial build for the uniform and non-uniform blanket cases. There are some possible changes in the minimum radius region. She is concerned about the possible neutron streaming in the helium access tube regions. She will assess the damage through the divertor region using 2-D or 3-D models.
She provided the background assumptions for the TBR assessment and then provided the overall TBR as a function of major radius and concluded 7.5 m would be the minimum major radius to achieve a TBR of 1.1 based on the 1-D analysis. A 3-D neutronics analysis will help to confirm the design basis. Changes in major radius could be coupled with changes to the lithium enrichment (nominally 90%) to meet the tritium breeding requirement.
Laila showed the calculated heat load to the various power core elements with an overall energy multiplication of 1.155. She also presented the power split between the He and LiPb coolants. She noted the machine design requirements are satisfied except that the divertor region needs to be checked and plant availability to be developed.
Laila presented the radwaste characteristics of the magnet structural materials: Incoloy-908 and JK2LB. The 94Nb is the main contributor to the Clearance Index value and dose after 1 year. The JK2LB is much superior to Incoloy in that regard. Leslie Bromberg should decide on the magnet material (action).
Status of CAD/MCNP 3-D Analysis - Paul Wilson reviewed the history of the CAD/MCNP(X) to perform 3-D neutronic analysis of complex configurations. The upgrade to Version 2.5.0 is complete with new added features and functionality. Other enhancements are in work. He showed his analyses on the ITER device. The ARIES-CS configuration has been modeled, but an incorrect n-source distribution was used. Now, enhanced geometry capabilities will provide better n-source definition and NWL at the first wall surface.
Some ARIES-CS CAD modeling issues should be avoided, such as overlapping volumes, non-mating surfaces, and mis-alignments of very complex surfaces. UW will resolve these issues with UCSD if detected in the CAD model.
Reactor Building Concept Assessment - - Les Waganer noted the need to develop a design approach and a conceptual design of a Reactor Building to develop a credible cost basis for the building. This depends on the power core size and configuration, maintenance features, and maintenance equipment. Based on the prior power core configuration, Les showed the general approach for port maintenance. It was noted that this approach needs to be consistent with the CAD definition, the cryostat was not shown, the central building core has been eliminated and the building top is now flat, not domed. Les will work with Xueren Wang to develop the building definition (Action Waganer, Wang).
Experimental Verification of Gas-Cooled T-Tube Divertor Performance - Said Abdel-Khalik provided an overview of the 2-D and 3-D analyses performed to date to assess the thermal performance of the helium-cooled T-Tube divertors. Numerical results indicate heat fluxes of up to 10 MW/m2 can be accommodated. The design is robust with respect to manufacturing tolerances and flow distribution. Extremely high heat transfer coefficients (>40 kW/m2 K) are predicted at stagnation points. Said reviewed the T-Tube geometry definition. He showed the calculated heat transfer coefficients. The test setup definition and geometry were provided. As the original tubes with the slots were fabricated, the slot width had significant width variation. As a result, a new brass tube was substituted and the slots were fabricated as a series of individual slots with small "bridges" remaining to stabilize the channel widths. Spacers were added between the tubes to ensure concentricity.
Air was used as the experimental coolant at both low and high pressure to simulate the higher pressure and temperature helium with good similitude. The predicted flow fields were shown that illustrated the flow dramatically decreased away from the inlet, while the temperatures increased from the inlet. The predicted and experimental results were well correlated at most locations. Far from the inlet, the correlation worsened and this is being investigated. Additional work will involve different gas flows and power levels.
Design Definition of Stellarator Coil Systems - Leslie Bromberg reviewed the superconducting options for ARIES-CS, namely High Tc (epitaxial on structure), Nb3Sn (wind and react), and Nb3Sn (new fabrication approach).
Leslie discussed methods of protecting the magnet from quench. He proposed a resistive quench with an internal energy dump. He showed several magnet design approaches for this protection scheme. He is developing a code to predict magnet performance with such an approach.
Coil Structural Design and Magnetic-Structural Analysis - Xueren Wang showed the coil structural design approach with CAD modeling. Incoloy 908 was the structural material assumed. Between coils, the inter-coil structure is nominally 35-cm thick and behind (away from plasma) the structure increases to 65 cm (strongback). Xueren used both shell and solid modeling approaches for stress and deflection analyses. The results were similar, but the shell solution is very fast and easy compared to the solid model solution. The stress and deflection were analyzed using the steady-state magnetic forces from the coils. High stresses were observed only in very small localized regions. Moderate stresses were found in reasonably small regions. Very small stresses were observed in large regions of both the intercoil and strongback structure. All stresses were below allowable values. Maximum displacements (deflections) were on the order of 2 cm in areas somewhat larger than the maximum stress regions. These deflections can be mitigated by either locally adding structural material or modifying the geometry to allow the coils to deflect into the correct steady state position when under EM loading.
It was decided that thicknesses of the intercoil and strongback structure should be thinned (or thickened) according to the local stresses and deflections based on the existing analysis; however, the analysis should not be rerun. Regions can be estimated by stress and/or deflection to determine thickness and percent of total coil structure surface area. These can be added to determine an average thickness for the intercoil and strongback (or alternatively, one single thickness for the entire structure). This will allow determination of a reasonable mass and cost for a tailored coil structure. Meanwhile, Les Waganer will investigate a low cost approach to fabricate the structure. (Actions for Wang and Waganer).
Updates of the ARIES-CS Power Core Configuration and Maintenance -Xueren Wang showed the radial build with the added local shielding needed to protect the vacuum vessel and the magnet against neutron streaming through He access tubes. He provided additional design detail in the port region. It was noted the port size of 4-m height is needed to extract the maximum blanket module vertical height of 3.9 m; hence, the transporter Waganer showed cannot enter the port and must remain outside the port. (Action Waganer).
Cross-sections were shown at other poloidal locations to explain module removal and replacement. Xueren explained the layout and sequence of removal of the non-uniform coolant access piping. Xueren described the dual cooled lithium lead blanket design approach with a hot skeleton ring for coolant plenum and piping and structure to connect to the cold vacuum vessel.
Support and Possible In-Situ Alignment of ARIES-CS Divertor Target Plates - The maintenance requirements for the divertor was described to define a better support and maintenance approach. The divertor module will be similar in size to a blanket module and would be exchanged through the same ports. At present, it is thought there would be 24 such divertor modules, with a surface area of approximately 3 m2. A divertor support solution must address the need to accurately position the plates relative to the magnetic field and not be unduly influenced by thermal expansion of the supporting structures. This probably eliminates mounting the divertor onto underlying blanket modules.
A support attachment method using the coolant piping was devised to solve the thermal expansion issue that has the added feature of position adjustment (in and out plus tilting in two planes). This provides the necessary degrees of freedom to position the divertor plates appropriately with respect to the plasma as well as being able to adjust for wear and erosion during the divertor lifetime. The thermal expansion of the coolant piping will be well characterized to allow proper positioning during operation. Remotely actuated adjusting features allow real-time adjustments. The underlying blanket and shielding are also supported on this structural support arrangement. A shield insert within the inner coolant tube minimizes neutron streaming down the tube (to be confirmed by Laila). It can be withdrawn for access of the internal tube cutting/welding machine during maintenance periods. The lifetime of the divertor module is planned to be identical to the basic blanket modules at 3.3 full power years.
Compact Stellarator Radwaste Management and Safety Analysis
Recommended Changes to ARIES Top-Level Radwaste Requirements - Laila El-Guebaly recommended the project revisit the 1997 top level radwaste requirements for no waste greater than Class C. She suggested avoiding the geological disposal and recommend recycling and clearance of the waste as opposed to low-level waste disposal in repositories. This approach would require removing during reprocessing some radionuclides that would be transmuted in special modules in each fusion reactor. The source material could be an alloying element, e.g., the Nb of the Incoloy-908 coil structure. However, a few team members did not think that isotopically separating the highly radioactive elements was the best approach. Perhaps a better approach is to isotopically modify the initial material usage at a cost to eliminate any materials that would convert to moderate and high level radwaste in fusion reactors.
Update of Pressurization Accidents in ARIES-CS - Brad Merrill described the details of the MELCOR computer model being developed for the ARIES-CS dual coolant blanket power core. One half of a field period is modeled. Brad presented the modeling assumptions and initial parameters. He showed the volumes and radial builds used in the simulation model. Given the initial conditions and possible failures, he showed the startup transient fluid and structure temperature predictions. The 10-3/yr failure scenario considered was a rupture of the single FW channel. The vacuum vessel pressures reach 2 atm and the rupture disk relieves to 1.5 atm. Brad concluded his talk with a list of possible failures to be analyzed.
Finalizing ARIES-CS Power Core Engineering - Rene Raffray recommended that the ARIES team quickly choose a final set of machine parameters as the final design baseline that will be used for the final analysis step. A major ratio of 7.75 meters seems to be a good compromise that will still provide a reasonable parameter space and retain all the desirable attributes. The divertor still is rather vague and needs to be quickly defined to provide a credible integrated design. Rene had a complete set of action items to be completed ASAP.
Rene mentioned that the 17th TOFE meeting will have an ARIES-CS session, so a good set of technical papers is needed.