ARIES-CS Project Meeting Minutes
14-15 June 2006
University of California - San Diego
Documented by L. Waganer
Ref: Agenda and Presentation Links: Project Meeting Agenda
Welcome - Rene Raffray welcomed the ARIES team again to UCSD. Thanks to the UCSD team for providing refreshments and the meeting room.
Status of ARIES Program – Farrokh Najmabadi reminded the team that DOE expects the design and analysis of the Compact Stellarator be concluded by September 2006 with all documentation completed by December 2006. Farrokh Najmabadi suggested that all documentation be delivered as technical journal papers in a Fusion Science and Technology Journal special issue. Additional shorter technical papers will be submitted to the 17th TOFE meeting in November 2006.
The date and location for the next ARIES meeting will be held in late September or early October 2006 at PPPL. A one-day summary/dry run project meeting will be held first. This will be followed by a one-day formal review meeting by invited reviewers. A small project team will make the summary presentations to the review team.
The need to conclude the study with a common design and analysis basis suggested a more frequent conference call schedule. The next telecons will be held on July 5, July 20, Aug 10, and Aug 24. Les Waganer will distribute the schedule.
Note: Action items follow the main text.
Compact Stellarator Reactor Integrated Systems Assessment
Final Radial Builds for LiPb/FS/He and LiPb/SiC Systems - Laila El-Guebaly reported that her radial builds were supplied to Jim Lyon's Systems Code, influencing its new results. She noted completion of last meeting's action items. There are additional items to be completed to finish the analysis for the compact stellarator. Laila reviewed the machine parameters and attributes that are common for both the FS and SiC machine concepts. The intent is to continue with the FS baseline and to assess the alternate higher performance and higher risk SiC machine.
The new baseline radius is now 7.75-m major radius. For this radius, Laila displayed the first wall and divertor surface contour map with the uniform blanket and divertor region (~76% area) and the non-uniform tapered blanket region (24% of the area). The new radial build and compositions for both regions were shown. The winding pack is increased in thickness from 18.1 cm to 19.4 cm. Les Waganer needs to change his coil cover thickness from 5 cm to 2 cm.
As alternate design and costing evaluations, Laila presented two concepts, both with comparable TBR. One is a uniform LiPb/FS/He blanket everywhere with no tapered blanket/shield. This requires a larger major radius for the power core. The second option (SiC structure) has a much higher heat transfer fluid temperature and corresponding gross thermal efficiency. Laila showed a list of significant differences. Additional factors not shown were the LSA=1, no He pumping power (~100 MWe) and deleted costs of the second primary He coolant system and its heat transfer system (~$150 M). Laila needs to provide a new tapered view for the LiPb/SiC system to Jim Lyon.Other miscellaneous tidbits: The 2-m thick bioshield is a part of the Reactor Building, but is not the exterior wall. There also may be a LiPb dump tank below the Reactor Building main floor. The circular superconducting PF coils will transmit minimal EM loads into the structure. Should the systems code eliminate the primary structure account ($82 M) since the 2 m bioshield extends below the torus? Spare lower PF coils will be provided and stored below the power core. The change-out guideline of the SiC in that SiC option is established by a 3% burn-up fraction
Systems Code Status - Jim Lyon reviewed the action items he completed including the neutron source profile, the delta-min thickness used in the code, and the power to the divertor. Jim summarized the systems code improvements since the last meeting. Lower neutron flux was used to revise the neutron and radiation flux to divertor. He also explained the redirection of particle flux in SOL and divertor region. Jim documented the power fractions to the various power core elements. He explained his approach to reduce the power and peak load on the divertor.
He accounted for 90% of electrical helium pumping power in the heat transfer system as useful recovered work. He changed the strong-back thickness from 35 cm to 28 cm per the new definition. The availability was increased from 76% to 85% in anticipation of Les Waganer's availability analysis. To determine the effect of the low cost coil structure lower cost results, Jim reduced the price of the coil structure from $56/kg to $20/kg. However, he should retain the $56 as the nominal cost of the structure for the baseline assessment. Deltamin was changed to be consistent with Laila's value. The final change was a parametric dependence of alpha particle loss fraction based on R, B, n, and T from Long Poe's calculation.
Jim discussed the consequences of the fixed alpha particle loss rate on the reactor model. Jim had removed the ECH heating option last meeting per a recommendation, but the use of RF heating had too many design constraints. It was decided to go back to ECH and representative cost data will be provided.
Compact Stellarator Physics Basis
Recent Developments in Plasma and Coil Configurations - Long Poe recapped the April action items and developments, namely improved alpha loss for the baseline configuration (N3ARE). He had also developed an A=4 configuration with lower alpha loss. There are potential solutions for further alpha loss reduction and improved flux surface quality by increasing the core transform.
Coils have been designed for the baseline configuration of A = 4.5 and R = 7.75 m with the PF coils added for startup and, potentially, plasma control. The cross section of the TF coils increased to a width of 74.3 cm and thickness of 19.4 cm.
Long Poe Ku also provided a cursory design for the A = 4 configuration. The coil outlines are virtually identical to the baseline set. It has similar effective ripple with a better alpha confinement characteristic. The coils are equally complex. The inboard winding pack needs some improvement.
He also showed further developments of configurations with high core transform, that is the hybrid SNS and N3ARE, which has a large core transform, low effective field ripples, good alpha confinement, excellent flux surface quality, decent inherent magnetic well, and good MHD stability properties. The effective ripples are very small and the alpha confinement is better than the baseline N3ARE. The hybrid can achieve stable kink modes at 5% beta.
An Update on Divertor Plate Geometry, Thermal and Alpha Heat Load Analysis - T.K. Mau explained the baseline R=7.75-m ARE divertor plate geometry is scaled from the 7.0 m case. He was able to generate closed flux surfaces inside the LCMS with field line tracing, and produce an ergodic region near the LCMS. This serves as the basis for the divertor heat load study. T.K. showed a HSR4/16 target baffle configuration and peak heating profiles on the divertor and baffle plates for three sections. A 3-D representation of the HSR4/16 divertor and target plate configuration was shown.
T.K. showed a tentative geometry for ARIES-CS based on the W7-X and HSR geometry. But the group thought the FW and LCMS offset at 5 cm is too small. The 5-cm SOL would be correct only at the closest approach, but it should really be 10-20 most places and 50 cm in the throat of the divertor. The CAD drawing shown on page 7 is incorrect in that the divertor should not conform to LCMS. It should be at a greater width from the LCMS near the divertor region and probably have an opening at the center for neutralized alpha particles to be exhausted. There should also be baffle plates shown. Also, there should be up/down symmetry with two divertor regions that span the cusp regions.
It was decided that T.K. should enlist the help of others to determine a solution. T.K.'s progress will be assessed at the next conference call in about 3 weeks to determine the next actions. The expected solution will be a viable divertor and baffle design approach, location of divertor and baffle, area coverage, particle and heat flux average and peak values, and heating distributions. Not all has to be accomplished within 3 weeks, but progress must be demonstrated to achieve a viable solution at the next meeting.
Compact Stellarator Reactor Engineering Assessment
Design Updates on the Power Core Configuration and Maintenance - Xueren Wang said that the Power Core design updates have been completed for the R=7.75 m baseline configuration and the drawings are, or will shortly, be posted on the ARIES web site. These include a new coil winding pack (74.3 cm x 19.4 cm) and maintenance port (1.8 m x 4.0 m) dimensions, new coil structure design, blanket module segmentations and component arrangement, cryostat and bioshield, and blanket module segmentation. The average intercoil dimension is 20 cm and the average strongback dimension is 28 cm. Xueren explained his process for determining the approximate coil structure thickness for acceptable stresses. The group recommended he make an assessment of the coil structure including large ports. These port openings will induce extra stresses that may require thickening the port structure. Xueren showed the ports he is currently considering. It was recommended he also add a port per field period for ECH. This port could also be used to assist the scheduled maintenance of the blankets and divertors.
The question of ECH versus RF heating was discussed at this point. RF is cheaper but the design and integration of the RF waveguides has not been included. Although ECH is more expensive, it was decided ECH should be the baseline heating and plasma initiation approach. Ports would be around 1 m2 for each of the three field periods.
The blankets are now segmented every 10º in the toroidal direction to accommodate the port width of 1.8 m. The blankets are segmented poloidally into 5 to 7 modules to accommodate the port height of 4 m depending on the plasma shape at that toroidal position.
The design of the bioshield was discussed. It is nominally 1.9 m in the vertical walls, maybe thicker for the floor below the power core, and maybe thinner on the removable ceiling if there is no manned access in that area during operation. The weight and pressure loading on the ceiling is a concern as represented. It was recommended a center support beam be considered to help support the ceiling. The ITER approach should be considered. It is possible the shielding could be segmented to ease installation. The shape may be changed to be a more dome-shaped structure, perhaps the upper part of a toroid, e.g., top of a bagel. A subteam will be assessing different alternate approaches for the cryostat and bioshield. A thin cryoshield thermally insulates the inner cryostat surface to cryogenic temperatures and maintain atmospheric pressure to a vacuum. There are significant pressure loads to react to the bioshield.
L. Waganer and X. Wang decided that the port blanket and shield modules needed to be nearly the size of the port opening to facilitate removal of other internally-removed blanket modules. Thus, Xueren increased the size of the port module. However this module is as large as two regular modules. The upper half is connected to the upper manifolds and the lower half is connected to the lower manifold. Xueren explained how the coolant tubes to the port modules would be disconnected and reconnected. Xueren further explained the process of removing the normal blanket modules by removing semi-circular shield rings to gain access for cutting and welding. This complexity had not been accounted for in Les Waganer's maintainability analysis.
An Approach for Low Cost Fabrication of the Coil Structure - Les Waganer reviewed the coil structure requirements and fabrication goals. The coil structure weighs approximately 3 Mkg (1 Mkg per field period). For the low cost construction, it is recommended each field period be a monolithic part and be fabricated on site near the power core. The proposed approach is to arc deposit the JK2LB steel material with CAM instructions to fabricate the part to a near net shape. Multiple deposition robot machines would be employed to achieve the desired build time. After a field period part is fabricated to near net shape, it will be moved into a second position for stress relieving with heating blankets and insulation bats. After stress relieving, it will be moved to the next location and machined with small robot milling machines operating on guide rails with accurate fiduciary datums. Les showed the typical cross section with the winding pack in the machined groove. Cooling can be provided in a number of methods. The arc deposition would deposit the melted layer like welding, with features being built up by starting and stopping the deposition of the melt layer. Overhanging features can be constructed with cooled slip plates. Milling the grooves and installing the superconducting cables can be accomplished with small robots guided by the guide rails.
The three parts would be built up sequentially in a temporary building just outside the power core building. The parts would be sequenced through deposition, stress relief, and final machining concluding with bringing the parts inside the power core. After all the larger parts are installed, the power core bioshield walls can be completed.
Les then presented the costing analysis methodology underlying the costing estimate. The cost of the JK2LB welding wire dominates the cost of the fabricated part. The process is highly automated with only labor for machine operators and quality inspectors.
Details of the Maintenance Approach - Les Waganer reviewed the adopted building and the maintenance approach. The pictures shown in this presentation is one generation old (R = 7.0 m). The graphics shown on the web site and Xueren's presentation are the current generation. Les showed a generic extractor machine in three sealed transfer chambers outside the power core bioshield ready to remove the bioshield door and port shielding. The next picture shows the extractor starting to remove the port shield and blanket. The pictures show them separate and smaller. However Les recommended a larger port module to facilitate removal of subsequent modules. A plan view shows the three transfer chambers for the removal of the blanket modules and transfer into a mobile airlock that will transport the blanket modules to the Hot Cell, not shown. This non-movable transfer chamber will allow very rapid removal and transfer of the blanket modules to the mobile transfer chambers.
Maintainability and Availability - Les explained the determining factors for the height and the width of the port and the blanket module size. He mentioned the need for a shield plug and its placement. He had previously mentioned to Xueren the need for a larger port blanket module. He then developed the logic for the removal of the first few blanket modules. Les showed a graphic on the sequence of removal of the blanket modules within a port region. The highest availability would gained by removing all the blankets at one time, every 3.9 FPY.Les went through the maintenance sequence with estimated times. He applied a factor of 4 to make the estimate more conservative. But the consensus was that he was too optimistic in his estimates. He will research on the time to complete remote maintenance. He noted the time to remove the local shielding to cut and reweld the helium coolant piping was not in the analysis and will be added. The unscheduled time is about 4 times the scheduled. It was recommended the ARIES-AT unscheduled time be adopted to be consistent with ARIES-AT. The power up times should be checked with fission plants. The cool-down times were about right.
Vacuum Pumping Systems Definition - Farrokh Najmabadi discussed the costs in the vacuum pumping systems costs. There is no documentation of these costs from prior ARIES studies. The costs are for the pumping system and the vacuum vessel costs. Les Waganer noted the price of the ARIES-AT vacuum vessel was $39.1 M in 2002$ obtained from a detailed cost assessment. The determining factor for the vacuum pumping cost is the base pressure. Mike Zarnstorff will obtain the NCSX base pressure requirement.
ARIES-CS Magnets - Leslie Bromberg discussed the choice of the Japanese steel, JK2LB, a low carbon, boron added steel with a thermal coefficient of expansion similar to the Nb3Sn superconductor. The conventional fabricated cost will be lower than with Incoloy, around $55/kg as compared to $60 for Incoloy for a welded structure. There will likely be a complexity factor added to the base cost estimating relationship. This is not the low cost approach shown by Les Waganer that will be an alternate fabrication approach.
Cooling the coils and structure are lower than previously estimated. Leslie said that removing the heat in higher temperature regions is more efficient. Leslie also discussed the modeling of the quench procedure. He is recommending periodic heaters to make the quench more controllable.
Update on Experimental Verification of Divertor Performance - Said Abdel-Khalik restated his analytical and experimental objectives for the helium cooled divertor. He summarized the T-tube divertor geometry and performance requirements. He reviewed the flexible experimental test setup. An intermittent slot helped stabilize the width of the slot. Air flows and pressures could mimic the helium operating conditions. He noted an enhanced analytical modeling of the pipe end conditions. There are several operating conditions that would yield significantly different thermal performance. The best thermal performance was obtained with a single parallel flow condition (flows in inlet and outlet pipes are in the same direction). He showed a design modification that would produce that flow arrangement for ARIES-CS.
ARIES-CS Power Core Engineering: Updating Power Flow, Blanket, and Divertor Parameters for New Reference Case - Rene Raffray showed a revised power flow diagram incorporating Jim Lyon's new data inputs. He then showed a chart that compared the fractional core radiation, edge radiation and divertor peaking factor for the maximum divertor heat flux of 10 MW/m2. He presented a table of power estimates for the baseline case. The performance parameters were shown for the Brayton power cycle as well as the performance factors for the dual cooled blanket. He discussed the divertor performance parameters. The divertor target heat load can be better accommodated by tailoring the plates for the several performance zones. It was recommended he adopt a Gaussian peaking factor of 10. Rene updated the power cycle data for inputs to the systems code.
Action Item List From June 2006 ARIES Meeting