Starlite Project Meeting Minutes
20-21 August 1996
University of Wisconsin


Attendees: C. Bathke, J. Blanchard, L. Bromberg, D. Cole, J. Crowell, D. 
Ehst, 
L. El-Guebaly, S. Jardin, H. Khater, J. Kulcinski, S. Malang, T. K. Mau, R. Miller, 
E. 
Mogahed, F. Najmabadi, J. Santarius, M. Sawan, I. Sviatoslavsky, D-K. Sze, B. Thayer 
(RPI), M. Tillack, L. Waganer, B. Wiffen, X. Wang, C. Wong

Attachments: (1) Action Item List, (2) Meeting agenda

Welcome and Opening

Laila El-Guebaly welcomed the Starlite Team to the University of Wisconsin facilities and to the "Best Place To Live In America." L. Waganer opened the meeting with a brief overview of the meeting agenda (see enclosure).

General DOE News

Bill Wiffen informed the group of the status of the DOE funding. The House has passed a bill with funding of $225M, whereas the Senate bill is at $240M. The Conference Committee is not expected to take action until after Labor Day. In regard to reorganization within DOE, there is to be a Science Division and a Technology Division. It is likely the Systems Studies program will be located in the Technology Division under the leadership of Sam Berk. Some programs are being directed with a team approach. If Starlite is one of these, then members of the DOE team may be drawn from both divisions.

Final Strawman Design Point

Chuck Bathke provided the group with a complete listing and results of his 17 August 1996 strawman. There are only a few small changes to the results between this set of results and those of the 6 May 1996 strawman. The plasma and first wall geometry had moderate changes; while the PF upper/lower and outer coils are smaller, have less current, and are relocated. Higher edge densities were proposed previously as a means to increase radiation to the FW and reduce the divertor heat loads. The effect has been incorporated in the calculation of current drive power (which increases as the edge density increases), and led to the adoption of a maximum edge desnsity of hs/h(0) ~ 0.2. The dependency of the Zeff on the current drive is also included. The power balance now includes the neon impurity and line radiation. On the engineering side, the inboard radial build is thinner by 4 cm; and the FW/B and reflector end of life fluences are now 15 and 45 MWy/m2, respectively. The maintenance ports are now included in the costing.

The system code parametric trade studies revealed a slightly higher COE (74 mill/kWh vs. 78 mill/kWh) at a higher plasma ion temperature. If the previous temperature had been retained, the CD power would have significantly increased. The total RF CD power remains nearly the same, but the distribution of the power among the subsystems changed. Still, the main cause of the 4 mill/kWh COE increase was the RF system cost increase.

It became evident that the PF coils at their current locations will require the higher field strength Nb3Sn conductors in place of the NbTi conductors. The specific value and the basis of the added cost for this higher field capability continued to be debated in the meeting with no resolution obtained. However, Ron Miller agreed to determine the value of the magnet unit costs to be used in the final (economic) strawman update [Action: Miller.] Leslie Bromberg suggested the change from NbTi to Nb3Sn should occur at a field of 5 to 6 Tesla rather than the previously thought limit of 8 to 8.5 Tesla. {PS. In subsequent discussions with M. Tillack, et.al., Leslie accepted the value of 8 Tesla as credible.}

Physics Analyses and Results

Steve Jardin summarized the ARIES-RS physics assumptions, analyses results, and recommendations. The main reasons for choosing the reverse shear plasma is the ability to achieve a Beta of greater than 4.9% and a BetaN of greater than 4.35 (90% of the maximum BetaN was chosen so that the likelihood of a disruption is extremely low). The physics basis was also chosen to deliver a bootstrap current fraction in excess of 90%. Accommodations to keep the surface heat flux in the divertor within acceptable limits have tended to degrade the CD efficiency. Specific CD efficiency effects are endemic to the ARIES-RS design, but the trends are generic to any reverse shear machine. Steve noted that it is now considered possible to reverse-engineer the plasma profiles to optimize for the best CD efficiency (as opposed to mapping out the parameter space of reasonably acceptable values of CD efficiency). But time and budget do not permit application of this technique to the current design.

The vertical stability coils have been designed and analyzed to stabilize the RS plasma profile. [Action: L. Bromberg agreed to calculate the ohmic heating of the vertical stability coils after S. Jardin provides the current and time history of the coils.]

TK Mau reviewed the recent RF CD physics analyses and results. The effect of changing the plasma impurity from vanadium to neon was small because only the ICRF is affected, and the power for this subsystem is small. The main contribution during this period was to define the CD scaling as a function of Zeff which is now used in the ARIES Systems Code.

Dave Ehst is concerned that TK's refueling repetition rate is 1/10th of his expected rate. [Action: Mau to verify the refueling (recycle) rate.] The group accepted TK’s recommendation on the pellet size and speed but was concerned about the local density profile change in the plasma edge. Ultimately, this density perturbation will need to be analyzed in detail. TK said it may be beneficial to shut off the RF power delivery system during the pellet injection for a period of 40 ms, but this action will not require shutting down the power supplies and RF sources and, thus, should not degrade the reliability of the RF system.

Dave Ehst noted that he is working more on issues than on design solutions. He has been concentrating on the sensitivity of the CD efficiency to the plasma profiles. He also discussed his analysis to compare several advanced tokamak physics regimes in order to understand the basic issues and to support the project choice of the reversed shear plasma. [Action: Mau and Ehst to complete their comparison/validation of the RIP/CURRAY code results.]

Economics

Ron Miller has continued to collect and analyze data on the electrical energy market and new financial approaches. He now has the ITER cost data and has begun to analyze it in order to properly apply it to the ARIES cost database. Currently, recosting the TF coils is a high priority item for him. The ALWR data also has been released by EPRI and shows their target cost basis being tied to competitive marketplace energy values (~ 43 mill/kWh).

Engineering Systems

Les Waganer presented the major maintenance requirements that influence the design of the power core building. These requirements include the need to contain nuclear contaminates, provide remote maintenance equipment to the maintenance port, work at several ports simultaneously, and access many (or all) ports. The basis for these specific requirements was to obtain high plant availability at low capital and operational costs. He reviewed the maintenance design approaches of S. Malang and then presented the concept of a single transporter that contains a maintenance machine to remove and replace a core sector. [Action: Waganer and Malang to decide upon a preferred maintenance arrangement.] Laila El-Guebaly noted that the primary biological shield is correct at 2.5 meters of concrete, but any tunnel or hot cell exposed to a removed sector should be 2.0 meters thick. Waganer also showed the ITER tranporter that housed the pellet injector.

Dick Cole presented his current configuration of the power core systems and maintenance provisions. It was noted in the plan view that the intersection of the removable sectors should occur at the outer perimeter of the blanket, not out as far as Dick showed. [Action: Cole to revert back to the scheme shown in the 6/22/96 meeting.] There was a lot of discussion about the vacuum duct size and location, but it was concluded that the present design should be sufficient. The most restrictive passage is in the upper area, and it could be moved closer to the TF coil case. [Action: Cole to construct a view (or views) showing the vacuum vessel duct routing and cross- section.] The shielding around the RF ports was correct in concept, but the thicknesses should be as defined by Laila either for individual or combined shields. More detail on the welded vacuum vessel seal was shown, but Dick should evaluate the suggested mechanical design suggested by Malang [Action: Cole]. The vertical stabilizing coils were shown, half within the shield and half within the vacuum vessel. It was suggested that the shield be full thickness and that the coils be attached to the back of the shield. The vacuum vessel should be bumped out locally to accommodate the stabilizing coils, if required [Action: Cole]. Igor Sviatoslavsky should confirm the design approach for the coil joints that connect the removable and the permanent elements and then transmit the design approach to Cole. [Action: Sviatoslavsky] [Action: Cole should show the thickness of the vacuum vessel port extension and the local thinning around the TF coils (if necessary)].

Jeff Crowell presented his work on the ARIES-RS EM disruption analysis. His current model incorporates a distributed plasma current and vertical displacement events (VDEs). The disruption scenarios considered were: (1) 10 ms current quench with a stationary plasma, (2) 100 ms current quench with vertical plasma drift, and (3) 2500 ms vertical drift followed by a 10 ms current quench. On the vacuum vessel, the maximum EM pressures were around 0.6 to 0.7 MPa, with stresses around 8 to 10 MPa for all cases. The stresses in the divertor plates showed more variability; but the maximum is still around 10 MPa, which is acceptable. The stabilizing shells need to be supported along their near-midplane edge.

Hesham Khater presented the activation results for the various Starlite materials irradiated for their anticipated lifetimes in the power core. Hesham then noted if these components met or did not meet the waste storage classification guidelines. If they did not, he noted the offending element. [Action: Bill Wiffen suggested Khater restructure his analysis and presentation to determine the impurity limits for each material so it can meet the waste classification guidelines. This would be of more benefit to the materials community to help guide materials development.] Bill questioned the accuracy of the material composition for the trace elements (impurities).

Elsayed Mogahed presented the results of his LOCA analyses that incorporated suggestions from the last meeting and conference calls. He analyzed four cases involving emissivity of 0.5 for all external surfaces and 0.8 for all internal coolant passages, an enhanced emissivity of 0.8 for all surfaces, tenelon as the reflecting filler, and vanadium as the reflecting filler. After some discussion, the group suggested adding the condition of cooling the low temperture shield to an initial condition of 150 deg C and then letting it rise after the LOCA commences. A second condition would be to maintain the low temperture shield at 150 deg C throughout the LOCA. Mogahed re-ran the code with these conditions and reported back later in the meeting. The enhanced emissivity and the cooled vacuum vessel helped lower the maximum inboard first wall temperatures, but they still remained around 1200 deg C (with tenelon). The maximum allowable temperature should be significantly lower to convincingly argue that either the module would not be harmed by the LOCA and could return to service (it would not leak but would have to be replaced.) [Action: Sze and Billone to document the maximum temperature excursion to allow structure to return to normal service (no damage) and the temperature to retain structural integrity (with subsequent replacement).] Several other fixes had been postulated, but the separate passive lithium coolant loop between the inner and outer first wall and the outer shield seemed to offer the best chance of significantly lowering the LOCA temperatures in the first walls. [Action: Sze should analyze the heat transfer capability and, if acceptable, size the system components.]

Bob Thayer presented his safety analysis results based on Mogahed’s LOCA (pre- meeting) results. Time-averaged, LOCA temperatures of the components were input into the model, along with the average plate thickness of each region. [Action: El- Guebaly and Thayer to update model to current average thicknesses of the design.] This analysis resulted in a dose release on the order of 1300 rem for tenelon and 1160 rem for vanadium (with no containment considered). The group questioned the application of the experimental model to this case. Concern centered on the mobilizing mechanism (air versus a vacuum) and the modeling of a flat plate versus a closed structure. The group was especially concerned about release of radioactivity from within closed volumes and thick plates. [Action: Thayer should summarize for review the basis for the experimental evidence and the application to the reactor model. D. Steiner should convene a review panel to examine the applicability of the model.]

Dai-Kai Sze explained that the CaO insulating coating on the inside of the lithium coolant passages remains the most attractive means of obtaining an acceptable level of pumping power with a liquid metal system. The testing of insulating coatings at ANL is continuing, and the results continue to look favorable. The present blanket design is capable of handling up to an average surface heat flux of 1.0 MW/m2 and a neutron flux of up to 5 MW/m2. [D-K, do we want to express this as a volume averaged number?] Dai-Kai summarized the scheme proposed by S. Malang to support the module sectors. Dai-Kai also mentioned the incorporation and the design approach of the vertical stabilizing coils on the back of the low- temperature shield. Bill Wiffen was interested in the technique of maintaining a separation of the lithium and the ferritic steel in the intermediate heat exchanger (IHX). Dai-Kai obtained a schematic of the IHX to illustrate the separation technique.

Dai-Kai presented Thanh Hua's thermal hydraulic analysis of the Li/V outboard blanket. Flow paths were shown for both the inboard and outboard blanket components. The incoming coolant temperature is 330 degC and the exit temperature is 610 deg C. Thanh Hua used a 3-D BKHEAT code. Due to toroidal independence, the analysis is reduced to 2D: poloidal and radial. He analyzed two cases: (1) poloidal variation with 1 MW/m2 peak at midplane, and (2) uniform 1 MW/m2 everywhere. The desired result was to obtain the required 610 deg C outlet temperature while not exceeding a FW peak temperature of 725 deg C — these objectives were achieved with acceptable coolant velocities. Dai-Kai also presented some vanadium alloy design issues from Mike Billone. Mike's recommendation on the operational upper temperature limit is 750 deg C for ductility loss due to grain growth and oxygen dissolution for Ti oxides. Lifetime issues were also presented (see viewgraphs for details).

Xuren Wang presented design details of the blanket with the schematic of the coolant flowpaths through the blanket and shield. [Wang - Need to send your viewgraphs to the principal organizations.]

Clement Wong presented a complete divertor design description and analysis that would meet the design requirements. He presented a figure which tried to describe the divertor power flow and the physics and engineering requirements. Refinements to the figure remain to be made. The key to finding a successful design envelope was the adoption of five key assumptions: (1) increase scrape-off layer (SOL) from 1 cm to 2 cm (at the midplane), (2) have an electron pressure balance in the SOL and divertor regions with equal electron and ion temperatures, (3) electron temperature decreases roughly linearly along a field line between the X-point and divertor strike point, (4) maximum particle heat flux < 4 MW/m2, and (5) the impurity enrichment factor (Ne for this case) at the divertor is adjusted to radiate the remaining transport power. For a Zeff of 1.7, the necessary enrichment factor is about 13, which is much higher that the experimentally observed value of about 3. The divertor surfaces are inclined to the separatrix by 15 deg outboard and 56 deg inboard. The total divertor area is now increased to 163 m2. The outboard and inboard vacuum slots are 6 cm wide. The maximum surface heat flux is now 5.8 MW/m2, with the average around 2.13 MW/m2. The next action is to determine if the thermal stress is within allowable limits or to suggest changes to bring stresses within allowable limits. The modeling of the power balance has been checked against experimental results, and it seems reasonable and consistent. Around 80% of the power is handled by the outboard divertor and 20% by the inboard divertor. The model assumed two toroidal ellipses in the divertor slots that radiate the power to the plates and the other surfaces. Modeling with 3-D elliptical cross-section toroids radiating to the complicated divertor and first wall surfaces proved to be a reasonably accurate approach.

Igor Sviatoslavsky presented his design for the divertor plates and supports, including techniques for manufacturing. He presented the cross-section and flat-pattern details of the 5-cm-thick plates. He discussed the problems in manufacturing the thick plate with coolant passages and the tungsten armor plates or tiles. Adjustable supports are to be thin-walled tubes with threaded ends to provide vertical adjustment. This design will be adequate for the anticipated disruption forces. The vacuum system design requirements were described, but it was noted that the external ducts need to be a certain size. [Action: The burn-up or recycle fraction values seem to be inconsistent. Bathke, Mau, Sviatoslavsky to reconcile.] Igor noted that the volume behind the outboard divertor plate is largely open space, but it will not be pumped. [Action: Wang should show the outboard plate extending to the support ring with no vacuum opening.] [Action: Wang - Definition of the divertor surfaces should be provided to D. Cole to incorporate in the overall cross-sectional view.]

Laila El-Guebaly presented the ARIES-RS radiation and shielding protection results. Hydrogen production within vanadium is predicted to be high, but acceptable. The higher values as compared with earlier estimates arise due to the use of the new FENDL-1 data. Use of this database yields a factor of 3 higher hydrogen production rate than using the older ENDF/B-V Xns database. Laila provided recommended shielding thicknesses for the local shielding around the RF systems that penetrate the power core shielding. Laila continued to recommend the use of a shield structure cost of $35/kg in place of the $68/kg to be consistent with ITER finding for their shielding cost. [Action: Miller to consider the ITER shielding cost value.] Laila also recommended 10-20 cm off-sets in the assembly gaps be included in the final drawings to reduce neutron streaming, [Action: Cole (and Wang?) include off-sets in shielding drawings.]

Leslie Bromberg summarized his PF and TF magnet designs. Some of the design details were from prior strawmen, but the general approach remains valid. [Action: Bromberg should update his FEM models to be consistent with either the 7 May or the 18 August strawman data.] Leslie noted that he would like to take the area on the side of the outboard TF leg for structure, which would help to thin the support structure immediately outside the outboard TF leg. [Action: Bromberg to verify the stored energy in the TF coil system which seemed to be too small.] There seemed to be some difference in the appearance of the TF coil support structure and the location of the PF coils in the 18 Aug (and also the 7 May ) strawman. [Action: Bromberg to update his FEM models to incorporate the current coil locations and forward the models to Cole for incorporation.]

TK Mau revised the powers for the current drive systems of ICRF Fast Wave, Lower Hybrid Wave, and High-Frequency Fast Wave. He felt that the changes in the power levels were consistent with a modest extrapolation of the system technology, but the group felt that the reduction of the ICRF power by a factor of 3 and an increase of the LHW by a factor of 4 required an update to the engineering definition. [Action: Mau to redefine the engineering parameters to be consistent with the RF CD subsystem powers. Provide new launcher sizes to Cole.]

T.K. Mau also updated the group on his recommendation on the 4-mm-diameter pellet to be injected at 5 km/s with a two-stage pneumatic injector. The physics group was still worried that the shallow penetration of the 5 km/s-pellet would cause local density disturbances. Higher speeds would be better, so the options of the rail gun and compact torus injection will be carried as possible system improvement options. Waganer and Mau did not know the method to guide the cryogenic pellet around gentle bends. [Action: Waganer and Mau to determine method to guide pellets.]

Programmatic

Farrokh Najmabadi outlined the necessary steps to timely conclude the study by the end of October 1996. He suggested that preliminary drafts of all sections be sent to UCSD by the end of September, which would allow review of the material and preparation of final drafts by the end of October. This would allow commencement of the next study in this time frame. All draft text and figures must be electronic (text in LaTeX and figures in PostScript). No scanned figures, use originals unless we need to refer to other published work and there is no other means of reproducing it electronically. Get approval release from author when referencing copyrighted materials.

Farrokh then reviewed the work needed to be completed in each section and the material to be documented in the final report. Most of these comments are as noted in the action items. We should make use of the ITER costing database where applicable for the ARIES-RS costing.

Farrokh said that the funding and budgets for next year are still undecided, pending the final DOE budget. It is anticipated that the study will be at a level lower than this current year. Several topics were reviewed for possible study. The next meeting was tentatively set for late October or mid-November at PPPL.



Action Items

Action Item List
Starlite Project Meeting
20-21 August 1996

Ron Miller Determine the value of the magnet unit costs to be used in the final (economic) strawman update.
S. Jardin Provides the current and time history of the vertical stability coils to L. Bromberg.
L. Bromberg Calculate the ohmic heating of the vertical stability coils after S. Jardin provides the current and time history of the coils.
T.K. Mau Verify the refueling (recycle) rate.
Mau and Ehst Complete comparison/validation of the RIP/CURRAY code results.
Waganer
and Malang
Decide upon a preferred maintenance arrangement for the power core building.
Cole Revert back to the scheme shown in the 6/22/96 meeting for the power core systems and maintenance configuration.
Cole Construct a view (or views) showing the vacuum vessel duct routing and cross-section.
Cole Evaluate Malang's suggested mechanical design on the welded vacuum vessel seal.
Cole The vertical stabilizing coils were shown, half within the shield and half within the vacuum vessel. It was suggested that the shield be full thickness and that the coils be attached to the back of the shield. The vacuum vessel should be bumped out locally to accommodate the stabilizing coils, if required
Sviatoslavsky Confirm the design approach for the coil joints that connect the removable and the permanent elements and then transmit the design approach to Cole.
Cole Show the thickness of the vacuum vessel port extension and the local thinning around the TF coils (if necessary).
Khater Restructure your analysis for the various Starlite materials to determine the impurity limits for each material so it can meet the waste classification guidelines.
Sze and Billone Document the maximum temperature excursion to allow structure to return to normal service (no damage) and the temperature to retain structural integrity (with subsequent replacement).
Sze Should analyze the heat transfer capability of passive coolant loop and, if acceptable, size the system components.
El-Guebaly and Thayer Update model to current average thicknesses of the design.
Thayer Summarize for review the basis for the experimental evidence and the application to the reactor model
D. Steiner Convene a review panel to examine the applicability of the model.
Wang Need to send your viewgraphs on the design details of the blanket with the schematic of the coolant flow paths through the blanket and shield to the principal organizations.
Wong Rework the figure which tried to describe the divertor power flow.
Bathke, Mau,
Sviatoslavsky
Reconcile the burn-up or recycle fraction values.
Wang Show the outboard plate extending to the support ring with no vacuum opening.
Wang Provide definition of the divertor surfaces to D. Cole to incorporate in the overall cross-sectional view.
Miller Recommend proposed shielding cost value.
Cole (and Wang?) Include off-sets that Laila recommended (10-20 cm) in final shielding drawings.
Bromberg Update FEM models to be consistent with either the 7 May or the 18 August strawman data.
Bromberg Verify the stored energy in the TF coil system which seems to be too small.
Bromberg Update FEM models to incorporate the current coil locations and forward the models to Cole for incorporation.
Mau Redefine the engineering parameters to be consistent with the RF CD subsystem powers. Provide new launcher sizes to Cole.
Waganer and Mau Determine method to guide the cryogenic pellet around gentle bends.