Starlite Project Meeting Minutes
June 20-21, 1996

(Held in Reno following the ANS Topical meeting)

Participants: C. Bathke, L. Bromberg, D. Cole, D. Ehst, L. El-Guebaly, S. Jardin, H. Khater, S. Malang, T.K. Mau, R. Miller, E. Mogahed, D. Steiner, I. Sviatoslavsky, D-K. Sze, B. Thayer (RPI), M. Tillack, L. Waganer, C. Wong

Guests: Dr. Ryoichi Kurihara (JAERI), Dr. Yashusi Seki (JAERI)

Attachment: Action Item List

Administrative

L. Waganer reviewed the agenda with minor revisions to accommodate ANS session conflicts and an expanded set of physics presentations. M. Tillack summarized the meeting objectives and the need to timely conclude the design and analysis activities and commence writing the final report. The intent is to publish the Starlite ARIES-RS by the end of the 1996 fiscal year. At the conclusion of the meeting, it was decided there are several systems that have not reached a mature and consistent design approach. These areas will delay establishing a firm final report deadline, but chapter authors should define topic, content, and authors for chapters within the next two weeks (Tillack agreed to collect and distribute all chapter outlines). The designated authors are to commence the final report writing immediately so drafts will be available by the next meeting.

A project conference call will be held on July 10th to finalize the technical approach and required performance on the undefined systems, notably the edge physics and the divertor design. (Note: Effective 3 July, the conference call number will be changed to 314-232-8169.) The next project meeting will be held at Madison, Wisconsin, 20-21 August. The following meeting will be at PPPL in late October or early November. The possibility of holding a community review (or some sort of advisory committee meeting) was discussed. If held in the August time-frame, the UW meeting should be extended one day. If held in conjunction with the following meeting, the value of a review should be closely evaluated in light of the progress in completing the ARIES-RS design and documentation. The possibility of holding a review in Washington, DC was discussed, but no conclusions were reached.

Preprints of all Starlite-related papers presented at the Reno ANS meeting should be sent to M. Tillack for collection and project documentation. Brief inputs to the 6th Quarterly report (including meeting presentations and five months of progress) should be sent to M. Tillack by 5 July. The Starlite Technical Assessment Report is nearly ready for publication with several sections available from the Starlite Web Home Page or archives. A potential conflict arose regarding D. Ehst's work reported in the Assessment Report and different results to be published in the near future.

Technical - General

In general, the 7 May strawman will be used for the technical basis for the final design. However, there will be another strawman run shortly to incorporate several necessary modifications (noted in the minutes.) Also there will be a final strawman used for documenting all final minor adjustments and reportable performance and economic parameters.

First Wall, Blanket, Power Conversion

D-K Sze presented a first wall blanket design to accommodate an average 1.0 MW/m2 surface heat flux and 4.4 MW/m2 average neutron wall load. At present, there is no peaking factor required for surface heat flux. Actual values may be lower due to changes in the scrape-off power balance. The final surface heat load requirements are to be resolved after the July 10th conference call. The blanket and shield design has been radially segmented into three zones (2.5 FPY, 7.5 FPY, and Life of Plant (LoP)) to maximize the useful life and minimize the waste and replacement costs. L. El-Guebaly reported the 2.5 FPY regions consist of the first wall, blanket, and divertor plates. The 7.5 FPY regions include the inboard replaceable shield, 38 cm of the FWB and reflector, and 20 cm of the divertor support structure. The LoP components are the shields, vacuum vessel, magnets, and cryostat.

D-K Sze presented a preliminary thermal analysis for the blanket and noted that T. Hua of ANL will do a 3-D thermal hydraulic analysis of the FWBS system. UCSD will do the 3-D heat transfer, and GA will do the 3-D coolant channel structural analysis. Thermal conversion efficiency of 46.4% was based on the advanced Rankine cycle data published by EPRI. The piping will transition from vanadium to 316 stainless steel upon exit from the vacuum vessel but the inside of the steel piping will be clad with vanadium to and from the sodium IHX/SG. D-K also reported the use of a Russian V-alloy tube with a new insulated coating of CaO. The percentage of Ca has been reduced, and the resultant coating integrity is much improved from this change. But MHD insulation performance will have to be demonstrated by experimental results.

There was some concern about the lack of design detail for the FWBS subsystem. It was reiterated that D-K Sze will complete the conceptual design, X. Wang will develop the CAD detail design, S. Malang will design the upper and lower supports, D. Cole will generate the overall configuration and gravity support, and M. Tillack will coordinate the design activity. The final shape of the FW baffle zone (transition to the divertor region) will be defined by Laila and Clement. Xueren Wang and Dick Cole will use this information, together with the inboard (IB) and outboard (OB) cross section drawings from Dai-Kai and radial build numbers from Laila to construct the CAD drawings. (See Divertor Section for discussion of new geometry changes.)

E. Mogahed reported results on his LOCA analyses. The ANSYS analysis was expanded to integrate both the IB and OB FWBS. Tenelon has been removed from the inboard replaceable shield as LOCA temperatures were excessive with afterheat contribution from Tenelon. Vanadium has been incorporated to lower the temperatures. Since the last meeting, the radiation and conduction FEM elements had been modified to better reflect the design approach used, but the team suggested the elimination of a 10% conduction term in the gap between the replacement layers. However, E. Mogahed suggested a design modification to retain the 10% conduction terms as it would mitigate an adverse LOCA temperature. His present analysis indicates inclusion of the conduction term would yield a maximum temperature of 1050íC IB and OB, whereas elimination of the conduction term would increase the maximum temperature by 100íC. A 10% conduction term was retained at the outside of the vacuum vessel to the support structure. [Action: Mogahed - Redo the analysis if design modifications for the additional conduction are not adopted.]

I. Sviatoslavsky presented his design for the vertical stabilizing coils which are located on the back of the low temperature shield. [Action: D. Cole - Incorporate stabilizing coils in the overall configuration drawings.] One-side cooling was adopted as the baseline approach. He also discussed the vacuum pumping system. The present space behind the shielding is very limited but D. Cole said he could modify the vacuum vessel to produce additional conductance area. [Action: Cole- Locally increase vacuum vessel to provide vacuum pumping space.] Turbopumps were determined to be inadequate; therefore, cryopumps were adopted as the baseline. There was some question about the quantity of particle flux handled. [Action: C. Wong - Compare particle flux and energy (power) out of the ARIES-RS and ITER divertors. Note: Wong and Sviatoslavsky have been working with ITER experts in this area, but it will continue to be coordinated and verified.] The divertor slot dimension was questioned, and it was confimed as 6 cm.

I. Sviatoslavsky also presented J. Blanchard's and Jeff Crowell's EM disruption analyses results. The scoping analysis assumed the loss of 10.4 MA plasma current with no plasma movement. It was felt this may be useful to address loads on the vacuum vessel but not on the first wall or divertor structure. A VDE condition would be a more severe condition. [ Action: J. Blanchard - Try to compare ARIES-RS to ITER's loss of current and VDE conditions for rough scaling implications.]

Shielding and Neutronics

L. El-Guebaly's FWBS segmentation scheme was noted earlier in the FWBS section. This approach saves 2000 tonnes of waste over the LoP and ~2 mill/kwh in replacement cost. She reported results of a 3-D neutron wall load analyzed with MCNP, V4A. The major changes were in the modified divertor configuration that reduced the local NWL at the inner plate from 0.5 MW/m2 to 0.035 MW/m2 due to the divertor throat area shadowing (shielding). Consequently, the inner TF magnets and the vacuum vessel are well shielded, however, the design will not be changed. These NWL values can now be scaled for similar geometies and more 3-D analyses are not needed. Some minor FWBS dimensions are to be revised in the next strawman due to the addition of the vanadium filler, 200 dpa V limit, gaps in the segmented blankets, heterogeneity of shield, revised magnets composition, and 40 FPY plant life. Shielding requirements were suggested for the current RF penetrations, which occur only in one sector. [Action: Cole - Incorporate shielding in the configuration drawings.] Note: All strawman inputs will be sent to C. Bathke on 2 July 1996.

Divertor Design and Analysis

C. Wong summarized the engineering aspects of the divertor. It had been assumed that, by radiation, most of the transport power would be uniformly distributed at the divertor surface. The peaking of 5 MW/m2 was calculated by considering the flux expansion of the SOL to the strike point, radiative power, and target geometry. This approach is supported by modeling and experimental results - similar approachs have been applied to ITER and PCAST. Part of the divertor radiated power will be directed towards the first wall and added on the core radiation. The total will be about 0.8-0.9 MW/m2, deposited fairly uniformly around the first wall. It was agreed in the last meeting that the first wall is to be designed to a conservative surface loading of 1 MW/m2. With these results, the divertor must handle a NWL of 1.1 MW/m2 (ave) and 2.5 MW/m2 (peak) as well as the surface heat flux of 1 MW/m2 (ave), 5 MW/m2 (peak).

C. Wong summarized Tom Petrie's results of his continuing parametric evaluation, based on the density and temperature inputs. Tom assumed Ar and Kr as the radiation species and determined the possibility of core and mantle radiation as a function of Zeff. With these assumptions, he found high mantle radiation power levels of 250-300 MW can only be achieved by higher edge density and lower edge temperature. T. Petrie found that only about 10 to 20 MW of power can be radiated from the SOL. These results are reported in more detail in the physics section.

After Steve Jardin presented his results of the Xe core radiation (see the physics section), the following course of action for the divertor design was adopted: S. Jardin will work with T. Petrie and T. K. Mau to include Kr radiation in the core, with variation in Zeff, and to recommend the maximum amount of power that can be radiated in the core and at the mantle while minimizing impact to the core performance. The divertor will take the rest of the transport power. X. Wang will evaluate the thermal hydraulics design impact with higher average divertor surface loading. C. Wong will go to the next level of accuracy to find ways of determining heat flux distribution in the divertor under the conditions of un-detached and partially detached plasma. This will have a strong impact on the peak-to-average ratio. At the same time, C. Bathke and D. Cole will consider the impact of extending the depth of the outboard divertor slot by about 0.5 m to increase the surface area of the divertor. There is no pumping port on the inner divertor slot, but perforations are permissible to pump from behind the plates. After the new slot depth is finalized, a new divertor surface geometry will be developed.

C. Wong presented some material erosion data from the DiMES experiment. Vanadium has roughly five times the erosion of tungsten. Considering the case of physical sputtering, tungsten was shown to be a suitable candidate if the high re-deposition rate can be verified. This is further supported by the operation of the radiative divertor where the ion temperature is in the range of 10 to 20 eV. This range is below the sputtering threshold of W and W may have a Maxwellian distribution of energy. C. Wong also identified other issues that will need to be resolved, including the attachment process and self-sputtering for the case of W-coating. Clement also presented E. Reis' 2-D mechanical stress analysis results. A flat 2-mm plate of tungsten was attached to the vanadium structure with integral coolant channels. The analysis allowed the geometry to expand but the edges were constrained to remain planar and parallel (i.e., no rotation). The design criteria was 3 Sm = 360 MPa. Preliminary results exceeded the design criteria, thus castellation and a thicker backplate will alleviate but not remove the high stress intensity. E. Reis suggested, and the team agreed, that constraint on infinite channel length without allowance for rotation was the problem. [Action: E. Reis - Investigate 3-D analysis and verify effects of boundary conditions.]

I. Sviatoslavsky updated the group on the mechanical design of the divertor panels. He showed a design and manufacturing plan to fabricate the tungsten-coated, cooled panels which C. Wong had described. The plate supports were sized and analyzed for static and dynamic mechanical loading. C. Wong reported the work done by X. Wang, S. Malang, M. Tillack, and C. Wong on the divertor surface definition, coolant routing, and thermal-hydraulics analysis. [Action: C. Wong - Generate a new set of TH and TS parameters for the divertor with the new heat load assumptions. Try to incorporate VDE-type loads into structure support scheme. Recommend W to V joining process.] [Action: D. Ehst - Provide thermal quench heat loads on divertor.]

Current Drive and Heating Design

TK Mau reviewed the current design approach and revised system parameters for the RF current drive and heating system. The ICRF fast wave power is 50 MW @ 95 MHz, the HF fast wave is 28 MW @ 1.0 GHz, and the LH wave power is 8.2 MW @ 4.6 GHz. (Note: These power and related launcher sizes will likely change due to changing Zeff to coordinate the core and divertor power distribution.) [Action: Mau - Recalculate system efficiency, power requirements, and launcher sizes.] Copper-cooled, vanadium structure will be used in all the launchers. Detailed cooling approaches were shown, which sparked the group to offer many suggestions. The general configuration can be incorporated into the overall design configuration. L. El-Guebaly suggested shielding recommendations for each type of launcher, the coax tubes, and an integrated RF module. Effects of irradiation on the RF cavity electrical losses were shown. Tillack and Wang offered to help TK define the cooling design and analysis. D-K Sze wanted to maximize the coolant and coolant temperature to maintain the overall power conversion efficiency.

Physics Analyses and Results

C. Wong presented a set of analytical data generated by T. Petrie. Based on the density and temperature inputs from D. Ehst (optimized for current drive) and T.K. Mau (favorable for mantle radiation), T. Petrie evaluated Ar and Kr as the radiation species, and determined the possibility of core and mantle radiation as a function of Zeff in the range of 1.7 to 2.5, and found that high mantle radiation in the range of 250 to 300 MW can only be achieved by higher edge density and lower edge temperature. Impact on bootstrap current and current drive has not been included in the evaluation.

Steve Jardin reported results from an analytical comparison of the physics models that included radiation terms. He used xenon as an impurity, varied the density profile, and obtained combined radiative powers for the core and mantle in the range of 99 to 409 MW. The flatter density profiles had detrimental consequences on the bootstrap current fit and the efficiency of the current drive subsystems. TK affirmed the decrease in CD efficiencies with increasing Zeff. D. Ehst suggested the use of a detached plasma to increase the heat flux to the general divertor region, but this was considered to be a last resort as it tends to move around unpredictably but may flatten the peak to average profile. S. Jardin stated that the core plasma would likely never radiate more than 300 MW and certainly never 400 MW. Realistically, the core might radiate 100 to 200 MW (including 80 MW of Bremsstrahlung radiation), but the divertor must handle the rest. Including Bremmsstrahlung radiation, C. Wong had been looking for about 300 MW radiation from the core and mantle. This discrepancy must be settled before or at the July 10 conference call.

Without making a presentation, S. Jardin showed C. Wong the work on the startup of the power plant, which shows the scenario of growing the plasma from the outboard, just below the mid-plane. C. Wong and D.K. Sze will have to consider the design impact of a limiter applied to the PFCs and on blanket modules.

Dave Ehst explained the analysis he has been working on to compare five equilibrium-stable plasma types, including reverse shear and second stability. It was reported that the second stability operating mode is thought to be less attractive than was previously reported by the ARIES team. Dave's ultimate goal is to author a refereed paper to Fusion Technology or Nuclear Fusion. His current findings are different than those reported in the unpublished Starlite Assessment report. [Action: Ehst - Resolve conflicts in the two reports in a timely manner so the assessment report can be published.] He compared the calculated results from Kessel, Bathke, and Ehst for stable plasmas. He suggested using a Zeff of 1.7 and a separatrix density of 0.6 x 1020, but it is difficult to maintain a high edge density with high frequency ELMs. His model now uses the full Hirshman-Sigmar model. The plasma needs to be fueled near the center, but it may be difficult for the pellet to reach the center of the plasma. [Action: Cole - Locate pellet injector for line-of-sight to center of plasma. Use a 10 cm dia hole or smaller away from the maintenance port.] The method of off-axis current drive is unknown. [Action: Ehst and Mau - Choose off-axis CD subsystem and give scaling algorithm to Bathke for ASC.]

Preliminary Safety Analysis

Bob Thayer did not have the promised safety results, but did describe the status of the safety hazard assessment. Steve Herring has been helping Bob a great deal with the formulation of the analysis and running the code. H. Khater suggested that a failure at the end of FW and blanket life was the worst for the waste condition, but another time would be worse for the safety analyses. He will supply Bob with the detailed conditions. The LOCA analysis will be used as the failure mode. Any new LOCA analysis data will be sent directly to Bob Thayer and Don Steiner. The assumption is that the worst condition will be that there will be no holdup in the release of the radioactive elements. The only limiting factors is the distance to the site boundary and the local weather conditions. The LOCA analysis will be conducted for the benefit of the Starlite program, but Bob's thesis research will have a broader scope. [Action: Thayer - Complete Safety Analysis for the LOCA conditions.]

The question arose about the correlation between LSA ratings and the system code economic modeling. We, as a project, have decided not to continue to use the LSA ratings to evaluate the waste and hazard potential, but we have retained the use of the LSA ratings in the algorithms to compute the plant costs. [Action: Miller and Steiner - Recommend a costing algorithm that is consistent with our safety assessement results (hence, eliminate the use of LSA rating).]

Magnet Final Design and Analysis

Leslie Bromberg described his current magnet configuration. For the March strawman, there wasn't a feasible poloidal field (PF) coil design solution. With the 7 May strawman, there is a PF coil solution, but it requires NbSn superconductor, which increases the PF systems cost by a factor of 4. [Action: C. Bathke - Investigate trade studies to lower PF coil fields to enable the use of the cheaper NbTi in the PF coils.] Leslie's TF coil design envelope still is shown as a trapezoid on the inboard leg which is translated to a similar trapezoid on the outer leg (largest points in toward plasma). This configuration increases the difficulty of extracting blanket segments. Leslie agreed the outer leg coil configuration could be converted to a rectangular configuration or an inverted trapazoid of constant area, or even the aspect ratio of the outer leg could be slightly modified for additional clearance. Excessive modification of the aspect ratio adversely affects the bending stresses in the outer coil legs.

Configuration and Maintenance

L. Waganer led the session with a summary of requirements to be met. He reviewed a table of maintenance time of the serial operations necessary to remove, inspect, and replace the power core components (not discussing parallel operations). The group began to critique the details, but decided to send all suggestions to Waganer for incorporation. [Action: Waganer - Incorporate all maintenance input and update in accordance with revised maintenance procedures.]

D. Cole of MDA summarized his understanding of the current strawman configuration and the general 3-D representation of the overall power core configuration. He illustrated the TF coil clearance schematic developed by Waganer a few months ago to help understand the segment removal clearances. Using this set of data produced an intersection of removable sectors inside the first wall boundary (not good!). Using the same set of clearances, but with an inverted TF coil cross-section, Dick achieved the intersection at the back of the blanket. [Action: Cole - Reduce the clearance sum to the 15 cm agreed upon during the last meeting and invert the TF coil cross-section (constant area). If the TF coils can be reduced in major diameter, Dick should inform Bathke and Bromberg ASAP so the PF coil configuration/field/cost can be reduced.]

Two vacuum door/structure concepts were presented. One used a vacuum door at a larger radius with an open structure to allow vacuum conductance behind the shield. This approach used an annular frame to attach the door and provide a vacuum seal. The frame approach was not adopted because the increase in the frame size exceeded the size limitation. The other design used a vacuum vessel door and structure as a common element that is closely fitted to the back of the shield. There is not enough space between the ports for adequate vacuum conductance unless added space is provided. The door is structurally secured and sealed with a welded wedge. Typical piping runs were shown for the removable and permanent components. D. Cole should work with Sze and Wang for requirements and coordination with the detail design approach. D. Cole also illustrated a transporter cask arrangement with a structural representation of a maintenance machine. The cask would be positioned outside the port and then transported back with a carousel to an airlock and onto the hotcell. The biological shield would be at the radius outside the carousel.

S. Malang presented two maintenance concepts with the biological shield immediately outside the cryostat. One concept used a large, movable flask that would service two adjacent ports and contain replacement modules and maintenance equipment. The second concept used a dedicated tunnel that would allow movement of the individual replaceable sectors and remote handling equipment. Questions about this concept were related to where the shield plugs would be stored and the size of the space to accommodate all the components and equipment. The contaminated surface is smaller for the flask than for the tunnel. There was no decision as to the better approach, which probably depends on the maintenance times.

S. Malang described the structural ring concept that assures a continuous structural element surrounding the outboard blanket, divertor, and the inboard blanket. Blanket elements will be structurally attached to the midplane point with a self-contained, retractable connector in the sidewall of the blanket. Similar attach connectors will be used above and below the midplane which will allow differential expansion of the blanket sector. A schematic of a coolant tube passing through the vacuum vessel was shown.

Systems Studies

C. Bathke reviewed the engineering inputs that were used to generate the 7 May strawman. The summary of the TF coil sizing trade study highlighted the influence of the TF coil size on the PF coil size and currents (and resultant COE changes). Chuck showed the new TF modified shape that is incorporated in the new strawman. C. Bathke will also have to increase the height of the the outer PF coils to accommodate the taller removable sector. The formulation of the CD system efficiency was described, which is roughly the average of the Mau and Ehst models. C. Bathke reported the Tion influence on the plasma core radiation fraction and the COE, both of which optimized roughly for the same Tion. - no change was recommended. A POPCON plot was used to determine the startup power requirements.


Summary of Action Items

Mogahed Redo the analysis if design modifications for the additional conduction are not adopted
Cole Incorporate stabilizing coils in the overall configuration drawings
Cole Locally increase vacuum vessel to provide vacuum pumping space
Wong Compare particle flux and energy (power) out of the ARIES-RS and ITER divertors.
Blanchard Try to compare ARIES-RS to ITER's loss of current and VDE conditions for rough scaling implications
Cole Incorporate shielding in the configuration drawings
Reis Investigate 3-D analysis and verify effects of boundary conditions
Wong Generate a new set of TH and TS parameters for the divertor with the new heat load assumptions. Try to incorporate VDE-type loads into structure support scheme. Recommend W to V joining process
Ehst Provide thermal quench heat loads on divertor
Mau Recalculate system efficiency, power requirements, and launcher sizes
Ehst Resolve conflicts in the two reports in a timely manner so the assessment report can be published
Cole Locate pellet injector for line-of-sight to center of plasma. Use a 10 cm dia hole or smaller away from the maintenance port
Ehst and Mau Choose off-axis CD subsystem and give scaling algorithm to Bathke for ASC
Thayer Complete Safety Analysis for the LOCA conditions
Miller and Steiner Recommend a costing algorithm that is consistent with our safety assessement results (hence, eliminate the use of LSA rating).
Bathke Investigate trade studies to lower PF coil fields to enable the use of the cheaper NbTi in the PF coils
Waganer Incorporate all maintenance input and update in accordance with revised maintenance procedures
Cole Reduce the clearance sum to the 15 cm agreed upon during the last meeting and invert the TF coil cross-section (constant area). If the TF coils can be reduced in major diameter, Dick should inform Bathke and Bromberg ASAP so the PF coil configuration/field/cost can be reduced.