ARIES Project Meeting Minutes
26-27 September 2012
Documented by L. Waganer
Ref: Agenda and Presentation Links: Meeting Agenda
Administrative and General Information
Welcome/Agenda - Les Waganer welcomed the ARIES team to the Bethesda Doubletree Hotel. He reviewed the day and a half agenda (see above link) and the topics to be presented.
Latest News from Fusion Energy Sciences - Al Opdenaker mentioned that the three grant proposals from the universities associated with the ARIES project are under peer review and are expected to be completed on time. The Continuing Resolution will affect the procurement process. The prospects for increasing funding in the years immediately following FY2013 do not look favorable.
Al mentioned that no results have come from the FESAC MFE Priorities panel thus far. The subcommittee is continuing to deal with the charge that was given to the FESAC on April 13, 2012. This charge can be found on the FES web page under the tab titled "Fusion Energy Sciences Advisory Committee." No date for completion of this charge has yet been established.
Status of FESAC Charge on Defining MFE Priorities - Dale Meade echoed Al's comments. He noted that only the subcommittee has met and later the full committee will address the charge. There will be no interim report, only the final report. There has been a lot of interest from the community with many white papers on the topic. The subcommittee has been reviewing the RENEW recommendations. Results from ARIES are viewed by the committee as being very informative and useful.
Status of ARIES-ACT Project and Near-Term Objectives - Farrokh Najmabadi was very pleased with the ARIES-ACT technical presentations at the recent TOFE conference. He asked that all TOFE papers should be forwarded to him so they can be added to the ARIES Publications web site.
Farrokh recently met with representatives from OMB who were interested knowing in the number of graduate students that the ARIES project has influenced and supported. They were concerned about the impact of ITER financing on the graduate programs. Farrokh asked all ARIES-associated university teams forward the number of graduate students influenced by the ARIES project. The OMB representatives were also interested in the cost of future power plants as compared to the cost of ITER and how long it would be before fusion would have a measurable contribution to the national electrical power production.
Farrokh stated that he would like to have all ARIES-ACT-1 definitions completed by December 2012 with papers published within another month. Then the team could complete the design definition and documentation of ACT-2 by the middle of 2013, pending suitable availability of funding. The remaining ACT-3 and ACT-4 efforts are anticipated to be only parametric studies with the ARIES Systems Code producing some comparative analysis.
Next Project Meeting and Call - The next meeting will be held in about three months, probably in the third week of January, 2013. Les Waganer will send out a Doodle query on the meeting date preferences for the weeks of 14-18 January and 21-25 January. The next ARIES conference call should be around the end of October or early November. It was later decided the call would be on Tuesday, 30 October 2012 at the usual time of 9:30 PDT.
ARIES Task Results (Physics)
ARIES ACT-1 Progress & ACT-2 - Chuck Kessel first presented his temperature and density profile variations for the previous smaller major radius of 5.5 m on the q-profile and MHD stability. These results will be compared with transport simulations from GA when they are available.
Chuck then presented the results of heating and current drive systems considering lower hybrid, ion cyclotron and electron cyclotron options.
Chuck reviewed the new 6.25-m major radius plasma equilibrium that features lower triangularity to obtain fewer horizontal inboard and vertical outboard flux lines. This larger plasma major radius produces new locations of the divertor surfaces (shown as superimposed on the viewgraph). The new configuration has longer slots on the inboard and outboard regions (OB 76 cm, IB 46 cm).
Chuck presented the results of several code runs for ACT-2 showing large power cores with disappointing parameters. Chuck mentioned that this is representative of using conservative values of all physics parameters and maybe this assumption is too conservative - perhaps some parameters might be adjusted away from the fully conservative case. Chuck then examined net electric power cases of 750 and 500 MWe to see if ACT-2 CD power and other physics parameters are better suited for lower net electric power values, but the Team was concerned that the economics of the smaller power plants would be very poor.
Chuck summarized his physics approach of generating an EQDSK solution for ACT-1 at 6.25 m. He realizes that all prior physics analyses were for the 5.5-m case and he needs to update the TSC simulations, H&CD analysis and ideal MHD solutions for the 6.25-m plasma case. He is not sure if the disruption analysis needs to be redone for the 6.25-m case. However a new PF coil solution will be required.
Chuck will begin the ACT-2 physics effort and expects the comparison of the ACT 2 physics to the systems solution to be valid. He anticipates the operating point to be at a large major radius, but it will be difficult to obtain the desired physics parameters and divertor heat flux at 1000 MWe. There will likely be a conflict between large fusion power and low recirculating power that requires high CD efficiency and low plasma current. The examined cases are likely to have high plasma elongations (>2). Only tungsten vertical stabilizing shells may be required in the outer blanket. The optimization quandary is that low betaN drives the power core to higher fields and high CD power.
Chuck identified several changes to the systems code that were recently imposed and are contributing to the degraded ACT-2 solutions as compared to Lane Carlson's earlier results of 2011. Those are the power scrape-off width, lower H&CD efficiencies, higher recirculating powers, lower thermal efficiency for DCLL blanket, and a thicker inboard radial build with the 75 cm thick DCLL blanket.
TAE-EP Interaction in ARIES ACT-1 - Katy Ghantous of PPPL was asked to examine the ACT 1 TAE-EP interactions. She explained the quasi-linear model assumptions and provided comparisons with DIII-D discharges where fast neutral beam particles are diffused. For one of the temperature and density profile combinations for ACT-1 at 5.5 m, she analyzed the linear stability with NOVA-K and found it to be unstable to TAE's, which are driven by the fast alpha particles from the fusion reactions. Then using the quasi-linear model, she predicted that 9% of the alpha particles would be lost from the plasma. Other plasma profile combinations will be examined, but any new analysis will alter the updated 6.25-m size.
ACT-1 Design Point Definition - Mark Tillack reviewed the approach for the design point selection and the progress made toward that goal since May 2012. The self-cooled lithium lead (SCLL) power core design is nearly complete. The "aggressive" physics database has been regenerated. Several new design changes have been made to the ASC, (namely, power scrape-off length, pumping power in the divertors, component efficiencies and geometric changes), which delayed code production runs approximately one month. Mark explained the details of the bucking cylinder thickness and BT max location changes.
Mark has been using the PPPL computer cluster to significantly shorten the code run time. A new set of output files entitled ACT-1d (R = 6.5 m) are posted on the ARIES web, ASC Results. This may not be the final set of design data for ACT-1.
Assessment of ACT-1d Cost Results / Cost Documentation Status - Les Waganer based his cost analysis on the ASC-1d's performance and cost code results of 26 July 2012. The cost analysis was intended to validate the incorporated costing algorithms provided by Waganer in 2009.
Les found a systemic error in all system and subsystem costs presented in the Costing Accounts.data file. The costs shown were about 9.5% high due to a misapplication of the cost escalation data. A few cost accounts have not been defined, but that needs to be done. There are other ASC output files, such as the Systems.data file that contains cost data that disagrees with the Cost Account file, perhaps by the 9.5% factor.
The naming convention or nomenclature of several systems and subsystems were not consistent and need to be corrected in the ASC. The High Temperature Shield and Support Structure needs to be classified as a replaceable item and noted that its operational life may be less than the life of plant because it is removed to the Hot Cell for refurbishment of life-limited blanket and divertor items. Laila El-Guebaly noted that if any component is replaced before its service lifetime, the impact on the radwaste volume and replacement cost should be included in the systems code. Radial builds quoted in the ASC data differ from those recommended by Laila El-Guebaly. Some of the discrepancies are due to the code calculating component thicknesses and others were errors. The materials database provided by L. Waganer was out of date and Waganer has provided an updated data set.
ARIES ACT-1 Nuclear Heating Profile (revised to reflect the latest changes to the VV design) - Laila El-Guebaly presented the final radial builds for ACT-1. She showed the ACT-1 CAD models (provided by Xueren Wang) that reflected these radial builds and contained more detail in the upper and lower divertor regions. These areas may be revised to some degree pending the location of the pumping ducts, He access pipes for the divertor plates, and the divertor configuration data from Tom Rognlien and Marv Rensink. Laila constructed a 3-D nuclear heating model of the ACT-1 SiC/LiPb-based power core. Laila noted that the nuclear heating results are scalable with the fusion power, which is provisionally set at 1804 MW for the 5.5 m major radius case analyzed (note that this plasma radius is being revised to 6.25 m with a corresponding 1814 MW of fusion power). Her results were shown for a 1/16 sector producing 102.3 MWth and the LT shield producing 4.65 MWth. She further described the heating inboard, outboard, upper divertor and lower divertor heating by component. She concluded the total heating is 1636 MW and the overall energy multiplication is 1.134. She noted that high nuclear heating (74 MW) is deposited in the LT shield (mainly in the inboard side), which will be dumped as low-grade heat. This prompted an action item to readjust the IB blanket thickness to capture most of this low-grade heat and enhance the power balance. The split of the heating between the high temperature coolants is provided (27% helium and 73% LiPb) yielding a total of 2054.64 MW of useful thermal energy including the surface heating.
Shielding Recommendations for ARIES ACT-1 and Activation of IB Components - Laila El-Guebaly reviewed the radial and vertical builds that were presented at the May 2012 meeting. She first showed the ACT-1 radiation limits with the goal of 1.05 TBR calculated and a net TBR of 1.01. The damage limits to the structure were presented along with the helium production limit. The limiting superconducting magnet parameters were shown. The ACT-1 (SCLL) IB, OB and vertical build thicknesses were provided.
Laila showed some options and rationale for the shielding of the maintenance ports, ranging from a thickness of 25 cm to 75 cm. The Team decided to have the LT helium-cooled shielding plug (55 cm) at the inner radius of the maintenance port. The vacuum boundary continues along the port walls and to the outer maintenance port door. The Team suggested there should be a tritium inventory analysis of the tritium accumulating on the walls of the maintenance ports.
Laila presented the waste disposal rating of the FS-based IB components using the commercially available common FS material specifications, which indicated that the Vacuum Vessel and Support Structure would produce high level waste products at end of life. If a more stringent set of material specifications, aka, "present" impurities, the WDRs of the components could be reduced to the category of low level waste. The cost of the upgraded FS material is not known, but Arthur Rowcliffe volunteered to see if he could find the approximate cost increase above common ferritic steels.
Laila provided the levels of IB and OB blanket decay heat as function of time for up to a year after shutdown. She concluded the LiPb, Support Structure and the Vacuum Vessel generate more decay heat than the SiC blanket structure. The tungsten stabilizing shells and the WC filler in the LT shield also generate high levels of decay heat. The worst case accident scenario would be the loss of flow accident (LOFA) in the blanket with a simultaneous LOCA of both the helium and water circuits.
3-D Tritium Breeding Results for ARIES ACT-2,-3 (DCLL) Blanket - Laila El-Guebaly explained the differences between the new 3-D TBR model with a thinner IB blanket as compared to the prior radial build model of 2011. The DAGMC neutronics code produces higher fidelity. The preliminary DCLL blanket on the IB and OB (no blanket in the divertor region) could yield a calculated TBR of 1.05 with a 90% Li-6 enrichment. These data are representative of a major plasma radius of 6 m and a minor radius of 1.5 m, which will certainly change pending more system optimization.
Laila explained the basis, assumptions and capabilities of the 3-D DAGMC neutronics modeling code, which is capable of doing detailed TBR modeling. To illustrate the impact of each step of the detailed analysis approach, she showed step-wise TBR results ranging from an infinite cylinder to the fully detailed sector with penetrations. She concluded by mentioning the key influencing TBR design factors that can increase or decrease the final results.
Updated ARIES ACT-1 Power Core Configuration and Maintenance for R=6.25 m - Xueren Wang showed the major physical parameters for the ACT-1(d) power core (6.25 m major radius with SCLL blankets). He then compared ARIES ACT-1d to ARIES-AT. ACT-1d has a larger major radius, similar blankets, slightly lower outlet coolant temperatures and thermal efficiency, helium cooled divertors and a vacuum vessel and LT shield outside the vacuum vessel. He showed a slightly updated, larger replacement sector, slightly tapered inboard blanket and longer divertor slots. The enlarged divertor slots (IB 48 cm, OB 76cm) may satisfy the edge plasma physics requirements.
Xueren provided detail on the integration of the helium cooled, tungsten-based divertor featuring both finger and plate designs depending on the final divertor heat flux and the design margin adopted
Xueren showed the overall layout of the power core components and how the maintenance can be accommodated. There remains some discussion about how the lower divertor region is vacuum pumped (or directed) from the bottom region up to the upper vacuum pumping ducts. It was suggested Xueren reassess if there might be sufficient space for dedicated vacuum pumping ducts at the bottom of the power core. There is a loose T-shaped attachment approach to allow positional adjustments for the sectors. These T-slots will be filled with a lower temperature alloy to anchor the sectors in place during operation, but they can be heated to allow installation and/or removal. Xueren provided a sequence of operations on how the sectors can be removed (and replaced). He provided a slide to stimulate discussion of the location of the maintenance port vacuum door. (Earlier in the meeting, it was decided that the door would be placed at the outer end of the maintenance power with shielding at the inner region of the maintenance port.) Xueren also asked about the method of vacuum pumping the lower divertor region - either use the large space between the Structural Support and the Vacuum Vessel or add a removable duct from the bottom divertor slot to the upper pumping duct. Earlier it was suggested that Xueren investigate making room for vacuum ducts at the bottom as this would enable more efficient pumping of the tritium in particular.
Initial Layout of the ARIES-ACT-2 DCLL Power Core - In order to help visualize the issues associated with the ACT-2 physical layout, Xueren Wang compared the ACT-1 (SCLL, advanced physics) and ACT-2 (DCLL, conservative physics) approaches. The final power core physics and geometry is not yet fixed, but it will likely be a larger power core with lower coolant temperatures and efficiencies. Although the blanket and first walls will be quite different, the divertors and vacuum vessels will be similar. Xueren discussed the design strategy for ACT-1 as compared to ARIES-RS and CS to frame the strategy for ACT-2. Xueren discussed the possible ACT-2 radial builds based on ACT-1 and older ARIES design strategies, but Laila El-Guebaly advised him to postpone working on ACT-2 until she defines a well optimized radial build.
Xueren has been considering the approaches for the cooling circuits to service the DCLL first wall and blanket subsystems. The manifolds will be behind the HT Shield/Structural Support. Xueren showed an example of the DCLL blanket design. The overall layout may look similar to the ACT-1 design.
Definition and Analysis of ACT-1 Vacuum Vessel - Farrokh Najmabadi described the analysis of the ACT vacuum vessel design. It will be a double-walled FS vessel, cooled with helium. The walls are 2.5-cm thick with an overall wall width of 10 cm everywhere. He provided stress analysis results for the most critical maintenance port side walls where the stresses peak. Vertical ribs will have less stress than horizontal ribs. The vertical rib case has lower stress values than the solid, 5-cm wall. The next activity is to analyze a full section of the port structure with the inboard ribbed structure.
Progress on Optimization Results for Steady-State Plasma Solutions - Alan Turnbull reviewed the progress up to June 2012 that achieves a successful plasma simulation using the Multi-Mode Gyro-fluid Transport Model (no plasma collapse). This model yielded two cases, but some issues remain. Alan believes the increase of the major radius to 6.25 meters will require creation of a new starting point equilibrium. Poor equilibrium convergence using TSC presents problems for transport evolution and reaching a steady-state solution. Employing lower hybrid and a small fast wave on axis produces suitable current drive, supplemented with the bootstrap current. As Alan integrated the transport solutions, the final profiles do not match the initial profiles. Chuck Kessel suggested Alan use his initial equilibrium solutions and PF coil parameters as they seem to be valid for steady-state solutions.
Small ELM and ELM-Free Operation in Reactor Conditions - Alan Turnbull proposed that small ELM and ELM-free plasma operation offers significant benefits if these conditions can be achieved and reliably sustained. He provided a few examples of each type of operation. He then described his proposed approach to assess the likelihood of achieving these beneficial small ELM and ELM-free operation. Alan noted that a new I-mode option is considered to be promising along with control of ELMS with paced pellet injection where small ELMs are triggered by each pellet. The I-mode can be described as having L-mode particle confinement, but with H-mode energy confinement. Simple scaling suggests I-mode can be scaled to an ITER plasma as well as a reactor plasma. The I-mode density profiles are similar to L-mode while its temperature profile is more similar to H- mode plasmas. The transition from L-mode to I-mode is proportional to density (ref. AUG and C-Mod). Alan still has some remaining questions (issues) regarding how small ELM and ELM-free regimes scale to the ACT-1 plasmas. Specifically, can they be present with higher betaN values? How does the high internal q with flat low shear profiles affect coupling of edge pedestal stability to the core and the plasma coupling to the wall?
Divertor Heat Load and Pumping Requirements for the ACT-1 Configuration - Over the past year, Tom Rognlien and Marv Rensink have been discussing fully-detached and partially-detached plasmas. Tom explained the differences of these two approaches with clear pictorials to help understand the interaction of the divertor surfaces, the plasma and the cold gas. Accompanying those pictorials, he presented some high level conclusions. The detached plasma has about half the divertor heat loads, however the particle throughput, helium pumping rate and neutral pressure still need more analysis. The slow evolution of the plasma and divertor models suggests feedback control might be beneficial.
For the new August 2012 magnetic equilibrium, Tom and Marv have found both detached and partially-detached divertor solutions. The fully detached divertor solution requires a wide slot and a flat bottom (separatrix perpendicular to plate) - a very important feature. On the other hand, strongly-tilted divertor plates are the key to achieving the partially detached plasma, e.g., ITER. The partially detached plasma used strong pumping on the inboard divertor region, whereas on the outboard region this solution needs strong puffing. The heat loads on the divertor plates are considerably higher than the detached plasma, but are still considered to be acceptable with advanced divertor design concepts.
Tom closed with high level conclusions for both types of plasma-divertor solutions. Heat loads are lower on the detached plasma. The particle throughput is controlled by pedestal transport for the detached case, but for the partially detached case, the throughput must balance reduced separatrix flux and reduced PF pumping. Helium pumping remains problematic and needs more investigation. The neutral pressures are hundreds of times larger than those for ITER.
Update on ELMs and Thermal Creep in the ARIES Divertor - Jake Blanchard discussed his analysis of the tungsten, helium-cooled flat plat divertor in the ARIES ACT-1 power core, see the ACT parameters assumed for large ELMs, small ELMs and old, small ELMs. Note that the new small ELM heat loads are about 50% higher than the old ones. The old small ELMs indicted a 20 μm melt depth, but the new small ELM analysis predicted a 100 μm melt depth. A melt calculation was not presented for the large ELM case.
Jake assumed the tungsten structure operated in the 1100-1300°C range and, in that range, tungsten does creep. However the conventional creep strain rate may not be applicable as the common tungsten creep rates relate to much higher temperatures - ITER is investigating the tungsten creep rate in our range of interest. Jake applied a curve fit of the creep rate data to analyze the mid-channel displacement (deformation). If the surface heat flux is reduced to 6.7 MW/m2, the surface temperatures can be in the range of 1400°C, but in the areas where creep can occur, the temperatures are in the 900-1000°C range. Jake illustrated both the armor thermal and pressure stresses. Jake then described the creep behavior (pressure, thermal and total) of the tungsten armor after 2 years of exposure. Jake noted that these results are highly sensitive to the creep rates that are not fully characterized in this temperature range. Jake investigated reducing the tile notch depth (increasing the effective wall thickness) by 1 mm. This change reduced the pressure stress while slightly increasing the thermal stress, yet the two-year creep strain is reduced by 23%. He also investigated a case where the notch is eliminated that alleviates stress concentrations, but increases the thermal stresses. For a 4-mm wall thickness at the center, the surface temperature is lowered by 162°C and the two-year creep strain is reduced by 15%. Lesser reductions were predicted for a 3.5 mm thickness.
Actions Necessary to Complete the ACT-1 Design - Farrokh Najmabadi concluded the meeting with an in-depth discussion of those team actions that are necessary to complete the ACT-1 design by December, 2012, with a full technical description documented by January 2013. These actions are recorded below.
Summary of ARIES-ACT Remaining Tasks