Starlite Project Meeting Minutes

2-3 August 1995 @ UCSD
L. Waganer with inputs from S. Jardin



These meetings served to select the significant physics and engineering features to be employed in the U.S. Demo plant conceptual design. The Engineering Group and Physics Group held pre-meetings on 1 August to obtain a group consensus of the major topics to be decided. Then the consensus results were reported by M. Tillack and S. Jardin in the main project meeting. Bill Dove briefed the project team about the general DOE budget reductions and the expected impact for the Systems Studies budgets for FY96 and beyond. F. Najmabadi solicited inputs regarding project direction and purpose. The group consensus was that: (1) the Starlite project will be responsive to needs of the national program and OFE will take the lead in providing the necessary input regarding critical issues (e.g., assessment of new plasma operating regimes and assessment of alternate applications of fusion power) and (2) the bulk of the activity should quickly focus on defining the most attractive tokamak demonstration power plant and completing the conceptual design within one year. This would help provide a focus for the overall DOE planning. If the design is sufficiently attractive, it may help stimulate additional congressional interest and increase future fusion funding.

Engineering Group Pre-Meeting

Les Waganer described the engineering considerations that must be considered to achieve the top level plant requirements. The key requirements relating to the first wall and blanket system were the use of commercially-relevant technologies, the generation of no waste greater than Class C, significantly better safety than present plants, and lower cost of electricity.

Mark Tillack described the European Union's (EU) blanket development program as being well funded, with plans to install two test blankets in ITER. He outlined their process to narrow the four current design concepts to a final two that are planned to be developed for testing in ITER. The requirements for the blankets are not intimately aligned with commercial aspirations, rather on more near-term goals such as material databases and existing fabrication infrastructure. As a result, all blankets were to use a variant of MANET ferritic steel as the structural material. L. Waganer echoed much of the EU thrusts from his recent Test Blanket Working Group (TBWG) meeting in Garching. Japan is also strongly pursuing blanket development along similar lines.

D-K Sze discussed the advantages and disadvantages of the lithium-cooled, vanadium structure blanket concept. The advantages are a simple design, low pressure, and efficient heat transfer capability. The concerns involve effectiveness and reliability of the MHD-reducing coating, safety of the lithium coolant, bi-metallic transfer in primary coolant loop (internally coated steel pipes?), and oxidization protection for vanadium piping.

Clement Wong presented the pros and cons for the helium-cooled liquid lithium breeder and vanadium-structure blanket. The advantage of the helium-cooled design is increased safety and a more familiar advanced heat transfer/thermal conversion system. Disadvantages include the need for high-pressure (18 MPa) and large number of helium tubes, and the necessary resolution of the compatibility issue between helium impurities and V-alloy.

Laila El-Guebaly did a preliminary cost comparison of the two designs and concluded that the (present) helium-cooled design was thicker due to the presence of helium as void and more costly due to the use of V-alloy for the thick coolant planum at the back of the blanket.

Igor Sviatoslavsky described the effects of disruption loads on the first wall and blanket structure. ITER is using a very conservative approach, but a demo or commercial plant will have a less-conservative design with a reduced frequency or intensity of disruptions due to a lower operating plasma current..

Having heard the pros and cons of the two competing approaches, the Group was polled to summarize their technical judgment. Both approaches would likely be feasible and both contained difficult and unknown areas. The conclusion was that the group endorsed the liquid lithium, self-cooled vanadium structure approach for the Demo blanket design. The following is a brief summary of comments:

* Li/V blanket design is perceived to be simpler than the helium-cooled approach, but more design details like the necessary structural support and the handling of disruption will need to be included in the Demo design.

* Unknowns involving compatibility of helium and V-alloy appear to be larger than those associated with the lithium bi-metallic effects.

* MHD coating for lithium coolant tubes seemed to be better understood but not in hand.

* The lithium design is perceived to be more elegant.

* Size, complexity, and cost of the recuperator and the closed gas cycle thermal conversion were raised as significant concerns.

* Helium approach offered growth to higher temperature and efficiency.

* Helium approach has apparently more industrial support that the lithium design.

* Safety concern was raised on the primary and secondary liquid metal loops of the lithium design.

* Low pressure and low pumping power for the lithium design was seen as an advantage.

* Lithium design is more along the current US development path, although utilities and nuclear system suppliers may be more comfortable with helium systems.

* Most engineering group members thought, on a relative scale, the lithium system would be safer (even the helium-cooled system has stagnant lithium).

* Lithium cooling is more fault tolerant.

Having settled on the Li/V design approach, D-K Sze and C. Wong outlined the tasks to complete the reference blanket and divertor/PFC designs. The first step is to define the configuration starting from the ARIES-II baseline but incorporating a more realistic blanket and shield support structure and vacuum vessel. This is to be consistent with a maintenance scheme to be defined. Disruption groundrules have yet to be established and scoping analyses completed. An approach for establishing intermediate design "reviews" was adopted to help keep the project on track and assure consistent technical baselines.

Physics Group Pre-Meeting

The purpose of the Physics Group meeting was to summarize the work and conclusions to date and to recommend a physics approach for the Demo power plant.

C. Bathke presented the results of the physics tradeoffs for Starlite Strawman. An improvement was made in the systems code in the calculation of the ratio of TauP*/TauE. Now there is a self-consistent calculation of both TauP/TauE (without the effect of recycling) and TauP*/TauE (with the effect of recycling). C. Kessel and G. Sager showed some data from existing machines that indicated Taup/TauE is in the range of one to four (TauP/TauE = 4 is recommended for DSC use).

Comparisons were presented between the Reversed Shear and the Second Stability designs. Plots were shown of COE vs ion temperature for the two designs, and it was noted that the point-model alpha-fraction results should be compared with the results from TK's more complete calculations. The variation with aspect ratio was shown, and it is clear that the minimum in COE will occur for aspect ratios between the two extremes of A=3 and A=4.5. So more work is needed for the intermediate aspect ratios of A=3.5 and A=4.0. The sensitivity of the COE to variations in the helium recycle coefficient was also examined.

D. Ehst explained his self-consistent current drive calculations for the Reversed Shear configurations. Detailed calculations of FWCD efficiency for the Reversed Shear Strawman were carried out for the A=3 and A=4.5 designs, including equilibrium, stability, and bootstrap current. For the A=4.5 design, the efficiencies were well correlated to the formula: gamma_B = a*Te**0.56 with a=0.38. This is also being used for the A=3.0 design with the coefficient a=0.18, based on a single unoptimized case. Note that the gamma_B is the bootstrap-assisted efficiency.

An optimization of relative COE with Te shows that an average Te of 10 to 13 keV is near optimum. This was presented on a Mass Power Density (MPD) vs Engineering Q (QE) value diagram.

C. Kessel showed results from his recent Reverse Shear Stability and Bootstrap Current Studies. Self-consistent stable reverse-shear equilibrium has now been calculated (with D. Ehst's help) for A=3.0 and A=4.5 at 90% of the maximum theoretical beta values. Sensitivity studies have been performed of the bootstrap current profile to slight variations in the temperature profile for fixed pressure profiles. It was found that for extreme variations in the temperature profile, the seed current would increase from 0.95 to 2.13 MA.

Current Drive requirements for the Reverse-Shear Demos were presented by T.K. Mau. The CURRAY Code has been improved to include fast alpha-particle distributions, High Frequency Fast Wave fast alpha, and thermal ion damping, and Kessel's bootstrap current package. The current drive requirements for the A=4.5 and A=3.0 Reversed Shear Demo have been calculated using LFFW(22-34 MHz) for on axis CD and HFFW(900 MHz) for off-axis CD. The total CD power differs from that output by the Demo Systems Code as being considerably lower: 45 MW vs 120 MW for A=3.0, and 50 MW vs 82 MW for A=4.5.

G. Sager discussed the Plasma Heat Exhaust Analysis for Demo Divertor and PFC Design. With no radiation, heat flux to the divertor plates in Demo would be about 80 MW/m2. This level needs to be reduced to approximately 1 MW/m2. It may be possible to obtain this lower level by a combination of effects, including: (a) selection of the proper impurity; e.g. C, Ne, Ar, Kr; (b) sufficient impurity concentration in the core, but compatible with fuel dilution; (c) sufficient impurity retention in the divertor; and (d) sufficient scrape off layer electron density. None of these can do it individually, but a combination probably can. The divertor must also be compatible with other physics requirements, such as MHD stability, current drive efficiency, confinement, and thermal stability. A package is now being prepared for the Demo Systems Code to calculate and monitor the divertor heat flux.

B. J. Lee discussed the factors that determine the dimensions of Slot Divertor for Demo. It is proposed that strong deuterium injection away from the divertor and impurity injection in the divertor be used as a way to concentrate high impurity levels in the divertor region. Some experiments done on DIII-D indicate that the impurities can be highly concentrated in the divertor by this technique, perhaps as much as a factor of 10, without substantially increasing the impurity concentration in the core. The gap between the plasma and the side wall of the slot for the reversed shear configuration should be bigger than 0.2 m to avoid being impacted by hot neutrals.

S. C. Jardin discussed the low-A stability and bootstrap current at low q* results that he and J. Menard (graduate student at PPPL) have been working on. They have examined some new cases with very low values of ITF/Ip (0.35) and found that high beta could be obtained, up to 80% stable beta, but that the bootstrap current was very low (15%) and the plasma current very high, up to 72 MA. These values would make a reactor prohibitively expensive.

It was decided by the group to recommend the use of the Reverse Shear plasma operating regime for the upcoming U.S. Demo power plant. The near-term work plan for the Starlite Physics group is as follows:

Reversed Shear


1. Develop stability limits for the A=3.0, 3.5, 4.0, 4.5 designs with and without off axis current drive.

2. Provide passive stabilizing plate and active current requirements to Engineering Group.

Ehst and Mau:

1. Compare FW central current drive calculations, choose the best frequency and spectrum, and calculate efficiencies.

2. Examine alternative off-axis CD techniques such as mode conversion, ECH, Alfven.

3. Calculate rotation induced by RF.

4. Check the alpha fraction and alpha-pressure calculations in DSC.

[Laila El-Guebaly asked to know the information regarding CD penetrations, dimensions, location, composition,...]

Lee and Sager:

1. Refine divertor impurity radiation modeling.

2. Prepare heat exhaust model and constraints for DSC.


1. Examine the effect of Zeff on the COE.


Bathke and Kessel:

1. Examine lower triangularity shape, especially stability limits.

Mau and Menard:

2. Perform current drive calculation at low A.

Figure of Merit Paper


1. Update individual sections


1. Take over editing

Commencement of the Starlite Project Meeting Presentations

S. Jardin summarized the results of the Physics Group pre-meeting (see prior discussion).

M. Tillack summarized the Engineering Group pre-meeting that selected the lithium-cooled vanadium structural breeder approach. The Starlite project is at a key milestone in the project plan. The results from the assessments are being completed and documented and the design phase has been initiated by the selection of the Demo engineering approach. Near-term (three months) and longer-term (one year) tasks were discussed by the Engineering group task area leaders, including FW/Blanket/Shield, Divertor and PFC, Magnet Systems, and Neutronics. During the next two weeks, the task area leaders will contact project members responsible for individual analyses and define a work plan within 1-2 weeks. The starting point for the engineering design is the ARIES- II fusion power core, using the latest strawman parameters from the systems code, ITER-EDA type of divertor design, based on the Reversed-Shear mode of plasma operation. The main thrust during the next three months will be in the following generic areas:

1. Conceptual Design Evolution

a. Assess design for key issues and requirements compatibility

b. Define additional engineering "boundary conditions"

* Disruption loads and design guidelines

* Overall maintenance[/configuration] approach [including the Vacuum Vessel and Structure]

* RF systems space requirements and radial build

c. Refine the ARIES-II configuration, particularly in thermomechanical configuration and mechanical interfaces

2. Interfaces to Other Task Areas

a. Provide a self-consistent engineering parameter list for the current strawman (required for the system code)

b. Obtain essential interface information from the Physics group:

* Conducting shells specification

* Divertor slot configuration

c. Provide engineering information for safety analyses

3. Preparation for more detailed design analyses

a. Neutronics

b. CAD

c. Electromagnetics

Following the preliminary definition of Demo, the second 3-month work period will focus on providing more detailed analyses and adding details to the conceptual design.

Mark Tillack also reviewed for the entire group the highlights and the selection process for the EU demonstration blanket.

C. Bathke summarized his strawman results for the entire project group. The results incorporated Laila's new shield definition and costing, which results in lower COE than previous runs. In his table comparing SS, RS, and LAR, he cautioned that the LAR values may appear lower than they should be because of incomplete engineering requirements/constraints. LSA parameters are still being used as cost-determining factors. RS is optimizing around 15 keV. However, it should be remembered that the RS case does not have all the engineering design features incorporated and does not represent a fully optimized case.

C. Bathke also mentioned his attempts to coordinate the Demo Systems Code (DSC) with a UCSD CAD model being developed by X. Wang. F. Najmabadi cautioned that, although some dimensional data would be derived from the DSC, the bulk of the (CAD) design information should flow from the Engineering Group.

R. Miller explained the importance of the discount rate on the fusion economics. He will be examining the unit costs and the general economic groundrules. He will also keep abreast of the availability analyses.

D. Steiner reviewed the progress on the licensing and safety efforts including the published and in-work reports. G. Cadwallader presented the results she and T. Dunn had accomplished on the preliminary hazards analysis for a test case involving the ARIES-IV design. Their analysis was a best estimate release fraction with no holdup in containment. A future analysis might consider the effect of some holdup in the containment. The governing guideline is the Protective Action Guide (PAG). The scenario assumed a primary coolant pipe break, external to the vessel, that allowed the breeder to heat up an additional 100deg.C and release (was it 400 grams?) of tritium. Questions from the team indicated that this accident of a primary line break would not release significant tritium, but perhaps other accidents could, such as a helium purge line break. But the analysis approach proved to be feasible and the data sparked interest in modifying designs to make the system safer. Cadwallader and Dunn agreed to make corrections in the input assumptions and rerun the case.

T.K. Mau discussed the heating and current drive system requirements and design approach. He is recommending the use of ICRF as the reference. We need to make sure we allocate sufficient space on the first wall for heating system components, such as antennas or ports. Fast Wave ICRF should be used as the reference system for the on-axis current drive system. He has not selected the off-axis CD system, which requires further evaluation.

L. El-Guebaly talked about the shielding system and the remote handling requirements. She was expressing concern that the ITER replacement times for the blankets and divertors were very long and would not translate into realistic maintenance schemes for Demo. But ITER was not designed for rapid maintenance and thus it will not validate a rapid maintenance scheme. Dose rates for the remote maintenance equipment that will work within the torus is very severe. Thus all such maintenance equipment must be hardened above most commercially available levels.

D-K Sze reviewed the engineering tasks to be accomplished (see engineering group meeting notes for a list).

L. Bromberg presented his results of failure analyses of the magnet systems. He used the ITER TF and TPX PF magnet sets as representative coil sets (mainly because engineering data was available to facilitate these analyses). Failure of a PF coil would not cause any problem. However a short in a TF coil would result in a stress level up to 900 MPa as the coil tried to become circular in shape. This condition would only occur if a short were suspected in a coil and the energy in all other coils was very rapidly dissipated. It would not occur if no action were taken. The main problem is in the lack of sophisticated sensors to detect coil failures.

C. Bathke discussed the Low-Aspect Ratio results he has obtained. At a meeting at PPPL, Martin Peng suggested some changes to help improve the modeling. C. Bathke has concluded that a shaped center post is required to obtain reasonable Joule losses in the centerpost. There has been a lot of effort to model a consistent plasma operating regime with current drive. There still remain engineering questions with the material properties in these high fluence regimes. Because of the severely declining budgets and the wish to divert as much effort as possible to the RS demo, this work should be economically completed and documented on a non-interference basis.

B. Dove addressed the group and explained the significance of the declining budgets. He presumed that, in light of the significantly decreased budgets being forecast, his budget for these systems studies would decline to the general level of $2 M. F. Najmabadi asked each team member to express his/her views on the best approach for the systems studies and for the fusion program in general. Overwhelmingly, the group favored continuation of the definition of a demonstration power plant to help establish program direction for DOE with fusion as the major trust of the system studies program. The team will be responsive to needs of the national program and OFE and will take the lead in providing the necessary input in critical issues (e.g., assessment of new plasma operating regimes and assessment of alternate applications of fusion power). We, as a project, should also be more proactive and visit DOE and Congress to stimulate more interest in the future of fusion.

Administrative - In order to help save budget, it was decided that the project meetings would be held four times a year rather than six. The next meeting is to be in San Diego on 25-26 October. An Engineering meeting may be held in conjunction with the early October IEEE meeting in Illinois. In preparation for the IEEE meeting, please circulate a draft of your papers by 15 September for project review. [Farrokh, please set up a distribution list for Engineering and Physics Papers.] The Engineering Group is to hold a conference call on Aug 16 [done] and the next project conference call is on 30 August. [P.S. The Physics group decided on a 29 August call.] The ferritic steel report is being edited and M. Billone is submitting the draft of his section. The Vanadium report still needs a section on materials and then it needs a thorough review. The second quarterly report is ready for distribution. The first quarterly is still in the review cycle. In order to speed up the process of writing and publishing the quarterly progress reports, it was decided to have the project meeting minutes serve as the quarterly progress reports.

Starlite Project Meeting Attendance List

2 -3 August 1995

Name , Affiliation

Bathke, Charles LANL
Billone, Michael C. ANL
Bromberg, Leslie MIT
Dean, Steve FPA
Dove, William DOE/OFE
Ehst, David ANL
El-Guebaly, Laila Univ of Wisc
Ellis, Bill Raytheon
Hofer, Gregory G. Raytheon
Jardin, Steve PPPL
Kessel, Charles PPPL
Lee, B.J. UCSD
Mau, T. K. UCSD
Miller, Ronald UCSD
Najmabadi, Farrokh UCSD
Sager, Glenn T. GA
Steiner, Don RPI
Sviatoslavsky, Igor N. Univ of Wisc
Sze, Dai-Kai ANL
Tillack, Mark UCSD
Waganer, Lester MDA
Wang, X. R. UCSD
Wong, Clement GA