Memo: Starlite-7

Date: 25 May 1995

Subject: Starlite Project Meeting Minutes, 11-12 May 1995 @ ANL

To: Starlite Team

CC: R. Conn, J. Coyle, J. Davis, J. Lang

From: L. Waganer + inputs from S. Jardin and M. Tillack (Thanks!)

Ref: Liquid Metal Insulator Workshop @ANL, 8-9 May 1995

Starlite Materials Town Meeting @ ANL, 10 May 1995

Attch: List of Attendees

Mike Billone welcomed the project team to ANL. Farrokh Najmabadi discussed the status of the project and illustrated the near-term activities and results on the project schedule. The decision on the main elements of Demo-I needs to be done shortly to accomplish a completed design by mid-1996. Bill Dove discussed the Congressional actions on budgets along with their possible impact. He stressed the need to report and publish Starlite results in a timely fashion. Demo is viewed as a very important project that will help shape the development of fusion. In addition to the all-electric fusion application, OFE is investigating alternative fusion applications. A splinter group of the Starlite Systems Studies Team may provide support on a non-interference basis with Demo.

Richard Mattas, of ANL, summarized the results of the Liquid Metal Insulator Workshop held earlier in the week. The Workshop had twenty-nine participants representing the U.S., Russian Federation, Germany, and the People's Republic of China. The objectives of the workshop was to update the status of research on liquid metal insulating coatings, review related blanket design, and identify key issues to be resolved. The new results are in that such insulating coatings can be successfully developed, but effort is required to demonstrate adequate performance in a fusion environment. Richard Mattas also reviewed the design status of the ITER Shielding Blanket. The current design is a modular shielding blanket, segmented both toroidally and poloidally. Approximately 800 of these modules are attached to a backplate that is permanently secured to the Vacuum Vessel. The method of module attachment, bolted or welded, is being carefully examined by the design teams. A separable first wall is to be provided. A decision on design approach is to be made in the Fall of 1995. The key issues are the severe environment, very large disruption forces and moments, stringent leakage requirements, and remote handling limitations

Les Waganer summarized the results of the ITER BPP Modular Shielding Blanket Cost Assessment Task. The design basis that was costed was the modular shielding blanket with an integral first wall. Most of the modules are welded to the backplate at the factory site and shipped to the ITER site for installation. In addition to the cost of the modules and backplates, the cost assessment included the costs for the in-vessel cooling system, the remote handling equipment, final assembly and testing, and related safety, QA and administrative activities. The dominate costs were the manufacturing engineering and the fabrication on the shield modules. Material costs for the modules and the cooling system and the fabrication costs for the first wall were also substantial. L. Waganer stressed that the costs represented work-in-progress and would not portray a final, optimized design.

Mike Billone reviewed the presentations given in the Starlite Materials Town Meeting from the previous day. The material data suppliers (material experts) offered three structural material choices with low activation properties; modified ferritic steel, vanadium, and silicon carbide. These are listed in order of decreasing availability of engineering data. The design community needs structural materials with known properties, high temperature operational capabilities, and a large manufacturing base. The exchange of information from the two groups benefited both in understanding their mutual requirements. M. Billone presented a preliminary list of material requirements, but the list needed some updating. Clement Wong and Dai-Kai Sze are to work with M. Billone to develop material standard performance requirements for Starlite.

Commencement of the Starlite Project Meeting Presentations

Safety and Licensing

Don Steiner presented the essence and progress of the Hazard Analysis tasks. First, the hazardous materials present in a commercial plant are identified (based on ARIES- II or -IV). The initiating events are being identified that could release a portion of the hazard source material. From these events, a sequence diagram is developed and the event sequence is quantified. The consequences will be determined for limited selected events and safety risks estimated. This is to establish the evaluation framework for the Demo designs. The Licensing and Safety group is pursuing a L&S workshop with the European Union (Ian Cook is the point of contact).

Greg Hofer outlined the tritium requirements that pertain to DOE and NRC governed facilities. Tritium is not classified as a Special Nuclear Material, thus there is no requirement based on existing laws. Also there are presently no NRC tritium safeguard requirements. DOE considers both gaseous and water vapor forms of tritium but NRC treats all tritium releases in a water vapor form. Greg was asked to find out more detail on how LANL TSTA is regulated for tritium inventories and releases.

Kathy McCarthy presented the status of the Fusion Safety Standards. Volume I (requirements) was sent from Ann Davies' office to the fusion community in early April 1995 with comments due back by the end of May. Volume II (guidance) is finished and is being reviewed at the present time. The Evaluation Guidelines for Public Exposure was presented to the team. A table of hazardous material comparison was discussed with comments that the choice and quantities of materials seemed inconsistent.

System Requirements and Modeling

Mark Tillack reviewed a proposed approach for evaluating the system requirements and translating them into subsystem design requirements. The remarks about the methodology were generally accepted, although a great deal of discussion over the requirements themselves took place. The requirement to adopt the same technologies to be used in the commercial plant met with disagreement and much discussion. This was one of the first requirements to be discussed, thus it is not clear to what extent this requirement stood out as compared to the others. For example, in previous meetings, there was similar controversy over the economic targets. Mark presented a draft of a table that allocates some of the high level requirements among the subsystems. Les Waganer later showed a similar table on allocation of RAM requirements.

Ron Miller discussed the economic modeling of the Demo and commercial power plants. He also examined the competition for base-loaded electric generating power plants, including advanced fission, natural gas plants, and advanced coal plants. The fossil plants of natural gas and coal are likely to experience a fuel cost that would escalate above the nominal inflation rate. This would suggest there is a window of opportunity for fusion if it's cost is sufficiently attractive to offset it's higher risk of a new technology.

Physics Overview

Steve Jardin presented a physics overview and discussed his Decision Logic Diagram. The physics group is now working in three areas: (1). Identifying top level requirements for physics, (2) Optimization and Self-Consistency of the 5 candidate designs (ARIES-I, ARIES-II, PULSAR, Rev. Shear, and Low-A) including equilibrium, stability, and current drive, and (3) Assessment of the credibility of the 5 designs in terms of physics parameters. [Note that the terminology of "designs" noted herein refer to the plasma operating regimes and the related engineering approaches represented by these referenced power plant designs and not the power plant designs per se.] Each of the 5 design concepts is being optimized separately. They are being assessed for physics credibility, technology credibility, and economics. From the top-level requirement activity, a selection criteria for both Demo-I and Demo-II will evolve. This will be used to evaluate the five (physics) designs and select one each for Demo-I and Demo-II. After these concepts are chosen, the Physics Group will define subtier level requirements for each.

T. K. Mau presented to the team the top-level requirements for the plasma-related subsystems including Heating and Current Drive, Fueling, Feedback Control and Diagnostics, and Power and Particle Exhaust. The Plasma System refers to the plasma inside the first wall and all the in-vessel subsystems required by the plasma to produce and maintain a flat-top condition in a safe and reliable manner. The plasma and the subsystems must be designed and function such that the related plant requirements can be met, specifically economics, maintainability, operability, reliability, safety, and waste disposal. They should be based upon developed technologies, capable of being extrapolated to levels required in a commercial device, and demonstrate operation in an integrated system.

Chuck Bathke discussed his results on the five physics strawmen analyses. He evaluated the Low Aspect Ratio (LAR) at an aspect ratio of 1.25 for 1000 MWe(net) to be similar to the other concepts. The results indicated a very high recirculating power fraction being consumed in the TF coils and Centerpost. Suggestions for improvement included decreasing the aspect ratio to 1.15, increasing the beta, and modifying the cooling model of the center post. The Reverse Shear device was modeled with some improvement obtained at lower aspect ratio (i.e., at A=3 from A=4.5). Modeling capability of the Reverse Shear at intermediate A was requested from the physics team. The requirement for minimal disruptions requires a 10% safety margin in normalized Beta ([[beta]]N). The credits for LSA should be removed from all comparisons. The Reverse Shear costs were improvements over ARIES-II second stability operation. Preliminary data on all five devices were presented to highlight differences and similarities. Chuck recommended several areas for improvement in the fidelity of the modeling.

Clement Wong expressed the intent to select the divertor physics and design approach to be employed in Demo by the next meeting. ITER is planning on using a radiation type divertor. The general operation and design requirements of such a divertor were explained to the team. The divertor surface material has not been picked but candidates are vanadium and copper. A surface coating of tungsten or beryllium may be employed. Glenn Sager described analytical models he has been working with to examine the relationship of the power exhaust and the PFC engineering parameters. It is anticipated the divertor heat loads may be considerably higher then those predicted for ITER. Minor impurity gases may be injected to enhance the radiation heat transport to the divertor.

C. Wong presented preliminary work on a helium-cooled first wall using a vanadium alloy as the structural material. A coating will still be required to protect the vanadium from corrosive impurities in the helium. To obtain high performance heat transfer capability, the pressure of the helium will be quite high (18 MPa). When coupled to a direct cycle gas turbine system and with te coolant outlet temperature of 650deg.C, a gross thermal efficiency of 45-46% can be projected. The team recognized the benefits of such a concept, but was concerned that this approach has not been examined in enough detail to adopt at this time for Demo-I. M. Tillack, C. Wong, and D.K. Sze will conduct a quick assessment of this approach in order to recommend an appropriate level of effort.

Laila El-Guebaly discussed the Tritium Breeding Ratio (TBR) analysis and breeding blanket material candidates for the Demo. Plots of data for ideal breeding blankets (no structure and no multiplier) were shown to illustrate the relative breeding capability of various liquid and solid breeding materials. A combined LiO2 and Li mixture was also illustrated. The change of the TBR for a range of content of various structural and neutron multipliers was shown. The model for tritium self-sufficiency was disputed in that it was not clear if some uncertainties were being double accounted. Laila discussed the top level requirements for the shield (bulk, penetration, and biological). Also reported was her success in optimizing the shield for a low cost design approach. By using the higher performance and cost materials in the inboard region and cheaper but efficient shielding materials in the less constrained outboard regions and divertors, the cost has been reduced from approximately $400 M to $200M, depending upon the ARIES design and materials selected. This corresponds to a reduction in COE of approximately 6 mills/kWh.

L. Waganer spoke about the need to design the Demo from the start with maintenance and availability requirements in mind. T. Flynn has stated that the BOP is expected to achieve 95% availability, thus the Fusion Plant Equipment availability must be 92% to reach the desired 87% requirement for the Demo plant. In order to assess the difficulty to reach that requirement, a table of elements that contribute to un-availability was presented to help define problem areas. Representatives of the Starlite Team will be asked to complete the table. Waganer was asked to provide a reference set of reliability data for pipes and joints as inputs for different components.

Tom Flynn presented some of European objectives (requirements?) for NET which were to achieve an lifetime availability of 10% and culminate at an availability of 25% for the last year. Estimates were made of the failure rates and mean downtimes for the components and subsystems. Failure modes, effects and criticality analyses (FMECA) with the initial data indicated the availability requirements could not be achieved. Components or systems were redesigned/improved or specifications were modified to achieve the desired results. ITER is also requiring periods of "high" availability during the Extended Performance Phase. Tom also showed a power train availability diagram for the Didcot B power train which calculated an availability of 94.2%. Data is also being retrieved from an archived RAM database for advanced nuclear components.

Leslie Bromberg followed up on his work on defining a maintainable magnet design concept. He briefly described prior fusion community efforts for a highly maintainable magnet system. ITER and TPX have also addressed maintainability of their magnet systems, but maintainability is secondary to initial assembly. Leslie is trying to examine the coupling of two coils to help react the overturning loads. He is also looking at using the shield to help support the out-of-plane loads. Leslie expects to make qualitative recommendations on the location of the cryogenic and vacuum boundaries in the near future.

Physics Figure-of-Merit Activity

Ehst presented his work on the Current Drive Figures of Merit. The basic current-drive figure of merit is proposed to be gamma_B = ne * I0 * R0 / PCD, where ne is the line-averaged electron density, I0 is the total plasma current (driven + bootstrap), R0 is the major radius, and PCD is the current drive power. Note that gamma B = gamma / (1-fB) for fB << 1. If R0, ne, and the fusion power are held fixed at 7m, 1.9E20, and 4230 MW respectively, then PCD scales like PCD = 615.7 / [gamma_B * sqrt(beta-N x q*)]. Note that the magnet variable costs scale with magnet energy, which scales like Utf = 138 GJ x (q* / beta-N). Varying q* at constant beta-N leads to an optimized value around q* = 3.84 which is near the ARIES-I value. Dave has reassessed the optimization of the ARIES designs with a simplified systems analysis code derived especially for the evaluation of the current drive systems. He has concluded that the prior design points are indeed the most optimal set of parameters. However, there was a lot of discussion that Dave should be working with C. Bathke to use and improve the ARIES systems code based upon the new analysis and results rather than generate and use a separate code.

Chuck Kessel described his Helium Ash Removal Figure of Merit. Removal of helium ash produced by fusion reactions is critical to sustain an ignited plasma, effecting both dilution of fuel ions and radiation loss. The following figures of merit were proposed to measure the effectiveness of ash removal: Tau-He*/Tau-E vs H/q* , Zeff, and some measure of the pumping speed. Here, Tau-He* is the helium particle confinement time divided by (1 - RHe) where RHe is the recycling fraction for Helium, and H is the energy confinement time enhancement factor over ITER-89P.

Reversed Shear Physics Progress

Chuck Kessel described a power plant-sized, reversed shear plasma that relies on the bootstrap current to properly position the q-profile minimum. The configuration relies only on fast wave current drive in the plasma center. The pressure profile has been broadened, relative to the TPX case, to position bootstrap current profile further out, which leads to a drop in beta-N from 4.8 to 4.5. The temperature profile has also been modified, relative to the TPX case, to obtain proper alignment of the bootstrap current profile and the desired current profile at the lower beta-N. Wall stabilization requirements are that the rotation be at 0.1 of the Alfven velocity (5.6E5 m/sec) and the resistive wall be within 1.3 a (1.72m) from the plasma center. The wall resistivity must be such that d/eta > 1.6E4, where d is the wall thickness. For vanadium, this corresponds to a thickness of about 1.5 cm.

Dave Ehst discussed the Reversed Shear current drive modeling. The preliminary results indicate a current drive power requirement of less than 60 MW of delivered power. The plasma is robust enough to allow a power disruption of 1 second or more with minimal effect.

Low Aspect Ratio Physics Discussions

T. K. Mau presented the RF Current-Drive Power Assessment for the Low-Aspect Ratio Demo. RF current drive power requirements have been assessed for stable, high-beta and low-beta LAR Demo equilibria. Low-beta cases are not of interest as Demo options according to the systems code analysis. For the high-beta case STREA06, a minimum of 143 MW of RF power is required, assuming no alpha damping. A low frequency (6 MHz) fast wave is needed for on-axis CD and a high frequency (250 MHz) fast wave for bulk CD. The RF CD power for the high-beta case is higher than assumed by ASC model, which is good only for on-axis FWCD. It is preferable to have a higher-beta equilibrium with good current profile alignment on the outer region and higher bootstrap current fraction.

Steve Jardin updated the team on Low-A assessment of the equilibria at an aspect ratio of R/a = A = 1.25. Several stable cases have been identified for further study. Cases that require some edge current drive are streac01 (47.7%,0.77), streac06 (42.9%,0.85). Cases that do not require edge current drive are streac08 (12.3%,0.85) and streac09 (11.6%,0.90). These cases all require a conducting wall at b/a=1.25 for low-n stability. A stable case without a conducting wall is nwhbsf32(27.7%,0.41). Numbers in parenthesis are maximum stable beta and bootstrap fraction.

Low Aspect Ratio Engineering Discussion

Mark Tillack reviewed the scope of the engineering assessment of the LAR tokamak. The assessment focused on issues related to establishing the main system parameters so that the system code could assess this option. Other critical issues related to the engineering systems design were noted but not pursued at this time. The status of the FS assessment write-up was reviewed. M. Billone will assume the final editing while M. Tillack is in Germany.

Clement Wong indicated that power recovery with He in the LAR divertor design is marginal. Cu-water probably works, but the heat is wasted. A large divertor chamber will also be needed in order to reduce the average heat load. Innovative solutions are needed for divertor heat removal.

Mike Billone recommended the DS copper for the LAR center post due to its mechanical properties. The conductivity drops significantly with Ni content. Creep and embrittlement are major problems. Water is probably the only acceptable coolant. In this case, corrosion is a serious concern.

Laila El-Guebaly discussed the LAR activation and shielding issues. For 3.6 MW/m2 peak inboard wall loading, the peak dpa rate for the unshielded center post reaches 50 dpa/FPY. The data on Cu shows rapid embrittlement and loss of work hardening by 0.1 dpa. For brittle Cu, rigorous stress analysis is needed to assure that all stresses (electromagnetic, pressure, bending, thermal, swelling,..etc.) do not exceed the stress limit for DS Cu. Increase of DS Cu resistivity due to transmutations will be reevaluated to include Co and Zn as well as Ni. Laila presented Khater's activation analysis which indicated that the center post will not qualify as Class C waste at any time of operation. In order to meet the Demo requirement of Class C waste based on both NRC and Fetter's limits, the center post should be shielded by 20 cm for 1 FPY (or 30 cm for 3 FPY) of operation. Adding 20-30 cm of shield on the inboard side would solve many radiation damage problems for the center post.

Leslie Bromberg reviewed the INESCO project results for applicability to the LAR concept. He also summarized LAR magnet issues, and proposed an innovative Cu-Li bed for the center post.

Administrative: Les Waganer will assume editing the first 1995 quarterly report inputs. Please send all inputs to him for inclusion. The next conference call is on May 31 (time of day is yet to be determined). The next project meeting is scheduled to follow the Utility Advisory Committee meeting on July 31-Aug 1 at UCSD. The Starlite meeting will commence on the morning of Aug 2 and proceed through noon, Friday, Aug 4. This is the meeting where the salient features of Demo-I will be defined, so please plan to attend the entire meeting.

Summary of Discussions and Recommendations for the First Demo Concept

Following the material workshop and the project technical discussions, the team discussed the project objectives, schedule, and general direction. The project objectives are to examine the existing physics and engineering databases along with a reasonable projection of R&D progress to be made in the near future, to determine a salable commercial embodiment of an electrical power generating plant, and to formulate developmental pathways from the next generation devices to the commercial plant. As a part of this, one or more conceptual designs of a demonstration plant will be proposed. The required R&D to support this demonstration plant will be identified and assessed. The established schedule is to develop the first Demo conceptual design in the mid-1995 to mid-1996 time frame, with a second design being accomplished the following year.

To date, the project approach was to utilize (adopt) the more well-developed conceptual commercial designs that are based upon more understood physics and engineering concepts for the first demonstration plant to be defined in the Starlite project. Several concerns or changes in attitude (Congressional budget pressure, enhanced customer expectations, and increased understanding of advanced physics and engineering materials/systems) prompted a re-evaluation of the project approach for the selection of the first Demo plasma operational regime and engineering systems. The team tentatively agreed to proceed to examine and propose a more agressive Demo design, balancing performance versus technical risk. Specifically, the Reverse Shear plasma operating mode would be proposed for along with vanadium as the fusion core structural material. The coolant is tentatively chosen to be liquid lithium, subject to the results of the vanadium/helium evaluation team. This decision on the design approach is not final until the next project meeting. Until that time, the Physics Group must conduct an in-depth examination of the Reverse Shear approach to determine that it would not impose an untenable approach.

The approach and decision for the embodiment of any other Demo design concept definition will be delayed until later in the project to allow maturity of some concepts, completion of testing and analyses for materials and components, and clarification of budget projections.

Starlite Project Meeting Attendance List - 11-12 May 1995

Name Affiliation

Bathke, Charles LANL
Billone, Michael C. ANL
Blanchard, Jake UW
Bromberg, Leslie MIT
Cheng, Edward TSI Research
Dean, Steve FPA
Dove, William DOE/OFE
Ehst, David ANL
El-Guebaly, Laila UW
Flynn, Tom Raytheon
Gomes, Itacil ANL
Hofer, Gregory G. Raytheon
Jardin, Steve PPPL
Johnson, Carl ANL
Karditsas, Nanos UKAEA (Culham)
Kessel, Charles PPPL
Malang, Siegfried FZK
Mattas, Richard ANL
Mau, T. K. UCSD
McCarthy, Kathy, INEL
Miller, Ronald UCSD
Najmabadi, Farrokh UCSD
Peng, Martin ORNL
Prokofiev, Yuro ANL/Efremov
Sager, Glenn T. GA
Smith, Dale ANL
Steiner, Don RPI
Sze, Dai-Kai ANL
Tillack, Mark UCSD
Waganer, Lester MDA
Wong, Clement GA