ARIES Project Meeting Minutes

17-18 September 1998 @ PPPL

Princeton Plasma Physics Laboratory

Prepared by L. Waganer

Attendees: M. Billone, L. Bromberg, S. Dean, B. Dove, D. Ehst, L. El-Guebaly, S. Jardin, C. Kessel, H. Khater, T.K. Mau, J. Menard, Ron Miller, E. Mogahed, F. Najmabadi, D. Petti (INEEL), P. Politer, W. Reiersen, D. Steiner, I. Sviatoslavsky, D-K. Sze, M. Tillack, L. Waganer, X. Wang

Guests: D. Meade (PPPL), M. Peng (PPPL), G. Voss (Culham)

Encl: Action Items, Meeting Agenda

Administrative and General Topics

Farrokh Najmabadi emphasized that the physics and engineering analyses are nearing completion and the design should be completed within the next few months. The next meeting is planned for 2-4 December 1998 in San Diego. The meeting will commence at 1:00 on Wednesday and conclude at noon on Friday.

It was suggested that the final design and analyses of ARIES-ST be documented in a special issue of Fusion Engineering Design Journal, similar to the approach followed by ARIES-RS. Task Area Leaders should plan their section topics and section content and begin to assign writing tasks. An outline for the FED issue will be presented at the December meeting (Action Item: Tillack and Jardin?). Suggested engineering papers for the ISFNT-5 symposium will be considered at the December meeting (abstract drafts are due 19 January 1998.) The IEEE symposium is also a possible forum for engineering papers.

Bill Dove discussed the likely outcome for ITER. The DOE fusion budget for System Studies is slightly increased for FY99 at $2.1M to $2.2M as compared to $1.4M for FY98. The ARIES systems study will share this budget with other studies such as advanced concepts, large non-electric plants, socio-economic studies, and use of magnetic materials in an MFE device.

B. Dove noted the many favorable comments and genuine usefulness to many researchers of the ARIES-RS design concept. It is being used as a near-term objective and metric for many new and redefined devices. Current experiments (e.g. DIII-D) are validating some typical ARIES-RS performance regimes. It might be feasible for our group to consider an advanced ARIES-RS study to extrapolate beyond the current experimental developments. A study of neutron source devices is also a possible area of investigation along with a new IFE study, but perhaps not for a year or so. First, we must complete the ARIES-ST design and documentation

Physics Assessment

NBI’s Role in Current Drive, Plasma Rotation, and Startup – TK Mau addressed the ability of neutral beam injection (NBI) systems to provide current drive, plasma rotation, and plasma startup up to full ARIES-ST plasma conditions. The ARIES-ST plasma is designed with a plasma equilibria having sufficient beta poloidal (bp) and optimized pressure profiles to yield a pressure-driven current fraction in excess of 99%. The remaining 1% on-axis seed current can be partially provided by a self-driven current arising from potato-like orbits near the axis. However, it is likely that all required plasma conditions may not be achievable with necessary precision. Some external control may be needed, such as providing current drive up to 5% of the plasma current.

TK noted that neutral beam injection is an efficient means of providing plasma current drive on axis. The beam will be injected near the plasma midplane. A 120 keV beam is adjusted to tangentially enter the plasma at an angle to obtain a driven-current profile peaked at a poloidal flux (y) value of 0.8. No attempt has been made to exactly match required current profiles. With the 8/25/98 strawman plasma parameters, a 5% current drive system would require 47 MW of power. Note that this set of parameters is for a design point different than the baseline strawman. TK's final viewgraph mentioned that the 8/25/98 strawman required 31 MW* of beam power. [*PS, TK mentioned that there was an error in interpolation and the correct strawman beam power should be around 44 MW.]

Neutral injection beams can also produce the desired plasma rotation as noted in JET and TFTR experiments. Moderate energy beams are more efficient in generating plasma rotation because of the high momentum content per unit power. Using the 8/25/98 strawman conditions and a 120 keV D+ beam, a plasma rotation of 66 km/s is required.

The NBI system can also provide plasma heating during the early plasma growth phases (before alpha heating is significant) and current ramping during the early and intermediate plasma phases. Some profile control may also be available. TK examined plasma shine-through and deposition profiles for three equilibria conditions provided by GA.

The proposed ion source and accelerator will be based upon the Common Long Pulse Source developed by LBNL, which is being used on DIII-D and TFTR. Eight ion sources would be required, combined into two beamlines. TK thought it might be possible to reduce the power required and use only one beamline, per the current configuration as shown by Wang. It is predicted the beam efficiency (Pinj/Psource) would be approximately 44%.

GA’s Physics Results Related to the ARIES-ST – Peter Politzer of GA examined his 0-D spreadsheet analysis results of the rampup of the plasma to the final size, shape, and current conditions (while maintaining axisymmetry) as a function of plasma current. The startup process commences with a small, circular plasma at the outer wall, increasing the plasma size (at constant elongation (k) and triangularity (d)) until (the circular) plasma is full width. Then the plasma is increased in elongation and triangularity to the final shape conditions.

Equilibrium plasma conditions (n=0) for the above analyses were verified with the TOQ code and the EFIT code. Self-consistent bootstrap currents were used with a Zeff of 2 and Lp/LT of 0.33.

Three schemes for noninductive plasma initiation were examined: helicity injection current drive, electron cyclotron resonance heating, and induction (using external coils). All approaches seemed feasible, but the electron cyclotron resonance heating was thought to best apply to the ARIES-ST concept, with a power requirement of 15-20 MW at 110 GHz. The plasma rampup primarily using bootstrap current and noninductive plasma initiation methods would be accomplished slowly over eight hours or so (1 kA/s). With the assumptions of ITER97H confinement scaling and a limit of 120 MW of external power, this time was affirmed. An improvement in plasma confinement would lower the power required.

Providing a higher order plasma shape factor (squareness) significantly improves the ballooning and low-n kink mode stabilities. Ballooning-optimized beta limit is increased from 58% to 67% by providing moderate positive outboard and inboard squareness conditions.

Peter estimated the heat flux present on the (inboard and outboard) divertor regions as 30 MW/m2. He mentioned that spreading some of the heat on the centerpost and contouring the divertor surfaces could reduce this level of heat. The Engineering Group opposed this, because the heat on the centerpost is already at critical design limits, whereas the divertor region is thought to be better able to handle the calculated heat loads.

The Dependence of Beta on the Aspect Ratios for ST Plasmas – Jon Menard reviewed our project’s current understanding of the ARIES-ST plasma equilibrium and stability considerations. Prior to the May 1998 project meeting, a plasma triangularity (d) of 0.57 appeared to be the maximum achievable. C. Kessel’s free-boundary calculations in that meeting showed that a triangularity of 0.64 is possible at low plasma inductance. By taking advantage of advanced kink stability over a range of toroidal mode numbers at higher plasma triangularity, operating at higher elongation (k) values, using profiles optimized for ballooning stability at higher triangularity (d = 0.64) values, enhancing the plasma parameters with the squareness profile enhancements, and relaxing the minimum wall separation distance to bwall/a = 1.15; a new set of optimized beta values were obtained over a range of aspect ratios at high bootstrap current fraction as shown below. Jon examined in detail the key plasma parameters for the strawman case with an aspect ratio case of 1.6.

Aspect Ratio

Elongation (k)















Equilibrium and Vertical Stability Analyses Results – Chuck Kessel explained that it is possible to have a low aspect ratio ARIES-ST plasma with high beta equilibria providing it has very low plasma inductance. He has generated free-boundary equilibria for the following ARIES-ST design point by prescribing plasma pressure and parallel current density profiles:

li = 0.125 b = 54% (60%-6% stability margin) kx = 3.4 dx = 0.65

Poloidal flux plots of this point were presented along with an optimized set of poloidal field (PF) coils. This optimized set of coils evolved from a large number of PF coils arranged on a rectangular boundary established to provide sufficient power coregeometry clearance. The engineering team suggested that the boundary could be moved to the outer surface of the TF return coil shell. Perhaps a few coils could be added or moved inside the TF return coil shell if they were normally conducting or heavily-shielded superconducting coils. If it would help in shaping the plasma or lowering the magnetic stored energy, the innermost coil could be moved closer to the plasma in a space behind the divertor shield. M. Peng would like to meet with C. Kessel to help understand the coil placement as he believes there are other coil arrangements that might be better. C. Kessel will also determine the coil design/arrangement that would create the desired plasma squareness.

Due to the weak gradient in poloidal flux on the inboard region, the inboard scrape-off layer is quite thick. However, the inboard first wall is only 5 cm from the plasma, which indicates that the power will be deposited over the entire length of the inboard first wall. From the latest strawman, 20% of the transport power is 69 MW and radiation power is 39 MW (guess). With an inboard first wall area of 105 m2, the surface heat load would be about 1 MW/m2.

In these low aspect ratio tokamak plasmas, conducting walls or partial shells are necessary to achieve vertical stability. An ideal MHD stability analysis using a perfectly conducting wall indicates the wall could be at 0.6 times the minor radius. A rigid vertical stability analysis with resistive structures (but with incorrect plasma parameters) indicates the wall could be at 0.53 times the minor radius. An ideal MHD stability analysis using a resistive wall indicates the wall could be at 0.6 times the minor radius. A fully nonlinear solution of resistive plasma and structure is in work, but numerical problems have limited the simulations. A conductive tungsten structure at the inboard divertor region closing off the inboard divertor leg seems to be required. No vertical stability control simulations with the TSC code have been successful.

Time-Scales for Current Rampup in STs and Tokamaks – Steve Jardin discussed the rampup times for the plasma current using non-inductive current drive systems. He estimated the time to slowly ramp the currents would be on the order of 1500 s.

Economics of Alpha Channeling in Tokamak Power Plants – Dave Ehst summarized the theory of alpha channeling and discussed his results on this subject. The theory of alpha channeling is to use radio frequency energy to enhance the alpha particle interaction with the plasma ions (not electrons) to help improve the plasma reactivity. Physics analyses have suggested that the alpha confinement times could be decoupled from the ion energy confinement times. The RS configuration would offer the best reactor embodiment. Approximately 50 MW of power would be required. However, D. Ehst concluded that alpha channeling would not be an effective solution. A member of the audience who was a strong supporter of alpha channeling kept offering alternative suggestions.

System Studies

Ron Miller summarized the ARIES-ST (1) systems activities to upgrade the systems code and background database, (2) technical presentations of results, and (3) new related socio-economic studies commencing.

Ron stated that he has been generating and posting on the ARIES web many incremental strawman cases for evaluation by the ARIES team. He highlighted the nominal case that the Physics Group has been discussing, namely A = 1.6, k = 3.4, d = 0.63, b = 0.55, fBS = 0.95 to 0.99. The results from the new strawmen are much improved from prior results. Ron acknowledged that his modeling of the EQDISK file produced roughly the correct plasma flux plots, with some obvious encroachment into the centerpost and an incorrect placement of the X-point. This was due in part to the routine used to place the PF coils works well with regular tokamaks, but poorly with low aspect ratio tokamaks. A revised EQDISK file from C. Kessel may help remedy the plotting routine. Ion density and ion temperature POPCON plots were shown but Ron was concerned about the validity of the plots. The COE has decreased somewhat due to the increased beta values of these cases. Ron presented results of several scans over engineering and physics parameter space. The COE was generally insensitive to plasma minor radius. A marginally better COE is possible with a smaller machine, but an improved design margin suggests the adoption of a slightly larger machine would be preferable.

It was decided at the end of the meeting to tentatively adopt the plasma and engineering baseline from the 8/25/98 strawman as the basis for the next and final iteration of ARIES-ST to be presented at the December meeting. Improved costing and system definition will be used to update the design concept at that time.

Engineering Analyses

Inboard Shielding Options and Impact on Design – Laila El-Guebaly presented results from her neutronic analyses based on the 7/24/98 Strawman – Case C (the newer strawmen cases show reduced power levels, but the results remain representative). Laila determined 15% water content for the 30 cm Cu shell and 15% borated water content for the 50 cm Al shell as an optimum coolant content to protect the superconducting PF coils. She optimized to coolant content in the shell and divertor shield to meet the radiation limits of the superconducting PF coils (same as ARIES-RS’). This aluminum shell will be 0.50 m thick (minimum) on the equatorial plane (thickness is proportional to 1/R at other elevations) to maintain constant conductance in the shell. [PS – This 50-cm thickness of aluminum may be increased if the TF shell cost can be significantly reduced as estimated by L. Waganer.] High performance tungsten carbide was recommended to shield the PF coils 1 and 2 (which are inside the TF shell) and to help reduce the centerpost height by 0.5 m. Optimum water content for these shields is 15%.

Laila listed the inboard shield design, safety, and economic requirements of the centerpost. A majority of these requirements favored a larger shield thickness, except reduction of the centerpost joule losses would favor a thinner shield (which allows a larger diameter centerpost and lower recirculating power.) The shield design concept, materials, and thickness will be an integrated, optimized solution for all key performance, safety, and economic parameters. Laila analyzed the impact of several inboard shielding materials (FS, W, WC, and WB) and coolants (He, H2O, and D2O) on the outboard breeding and radiation damage to the center post. Among all options, the FS/He shield offers the highest breeding and enhances the overall power balance.

L. El-Guebaly and R. Miller will conduct an integrated trade study of inboard shield design concepts versus recovery of useful heat.

Laila discussed how the low aspect ratio geometry of a generic spherical tokamak and the specific ARIES-ST blanket material choices have reduced the tritium breeding ratio (TBR) to a marginal level of 1.1. The 10% margin is provided to account for uncertainties in cross-section data for lithium and lead, TBR modeling, tritium losses, and export of tritium for new plants.

Shields for ducts and penetrations need to be defined. Definition of the duct locations and geometries are to be defined by Mau (NBI), Wang (piping, vacuum), and others.

Inboard Shielding Design – Igor Sviatoslavsky discussed the boundary conditions for the inboard shield and first wall. Igor presented an initial design for the inboard shield and first wall, discussed its advantages and disadvantages, andthen followed with a new design having improved shielding, thermal performance, and fabrication attributes. Igor (and Dai-Kai) suggested increasing the helium pressure to 12 MPa for the inboard shield to help enhance the thermal performance of the shield and other cooled components. He concluded that once-through helium cooling of the inboard shield and first wall can keep the maximum structural temperature below 500°C. No heat transfer coefficient enhancement techniques will be required. The electrical pumping power will be 11.6% of shield and first wall thermal power removed.

Blanket and Power Conversion System Summary – Dai-Kai Sze noted that most of the blanket, power conversion, and tritium recovery system analyses have been completed. He suggested the helium coolant pressure be increased to 12 MPa to improve the heat removal efficiency. Farrokh Najmabadi was concerned that there is too much power being removed by the helium coolant (2080 MW for helium and 1360 for LiPb.) [PS. Laila thinks the power split should be around 50/50.] This level of power being removed by the helium causes a pinch-point problem in the intermediate heat exchanger. Dai-Kai will confirm the power split and develop a new heat removal system to eliminate the pinch-point problem. This might involve elimination of the intermediate heat exchanger. Dai-Kai thought that the 46% cycle efficiency our team has been using may be too high for the coolant conditions that are evolving – 41% might be more realistic.

The ARIES team has been using the guideline that the daily tritium release to the atmosphere is 10 Ci/d. Dave Petti advised the team that we should strive to be as low as 1 Ci/d although this may not be achievable. Dai-Kai discussed existing permeation windows for both the LiPb and the helium coolant streams, which might improve the tritium coolant recovery capability.

The present approach to achieve a high temperature coolant (and high cycle efficiency) is to use a layer of SiC as efficient thermal insulator between the LiPb and the ferritic steel blanket structure. It was thought that SiC was not affected by long exposure to high temperature LiPb. But Dai-Kai showed photomicrographs and samples of SiC after exposure to high temperature LiPb that indicted SiC fiber damage. He noted that the SiC material was supplied by McDonnell Douglas (IMPORT tubes) some time ago and the material is probably not representative of current materials.

Dai-Kai mentioned a new breeding material, LiSn, that has low vapor pressures on the phase diagram and may represent an improvement over the current LiPb.

Divertor and Blanket Thermomechanical Design and Analysis – Mark Tillack emphasized that all of the Engineering Group should be using the web-based Design Data Book as a source of common, baseline data. All existing and new data should be put into this database as soon as possible.

Mark noted that previous thermal-hydraulic analyses indicated that a surface heat loading of 0.9 MW/m2 is possible with a 3-mm first wall thickness, roughening one side of the coolant channel, and allowing a 600°C structural temperature. Recent thermal-hydraulic analysis results showed that a value of 10% pumping power (electrical pumping power/thermal power handled) is adequate and 5% would be reasonable. Mark referenced the EU blanket design which meets the code limits with a 0.5 MW/m2 surface heat load; but we must evaluate our design with a 0.9 MW/m2 surface heat load (actually used 0.95 MW/m2). A finite element model (FEM) was developed with one and one-half coolant channels, which represented the calculational limit of the analysis code. The ARIES-ST outboard first wall is within all stress allowables at a surface heat load of 0.95 MW/m2). Thermal stresses are limiting, whereas pressure stresses are not. A schematic of the first wall coolant routing was shown.

Mark showed the current divertor cross-section with the large outboard slot divertor and no inboard divertor. The inboard shield with a tungsten surface intercepts the inboard flux line. The team suggested modifying the upper outboard divertor surface to increase the vacuum conductance to the vacuum pumping duct and decrease the neutral gas density in the divertor region. Tungsten will also be used on the divertor plates to accommodate the design surface heat loads of 3 MW/m2 average and 5 MW/m2 peak surface heat loads. Both embedded tungsten rod and slotted duct/brush approaches were considered but the fabrication and reliability of the slotted duct/brush looked more reasonable. Both thermal-hydraulic and thermomechanical analyses were conducted on the slotted duct/brush design. Reasonable slot dimensions and coolant velocities yielded acceptable pumping power and structural temperatures and stresses. The porous heat exchanger was also modeled. The analysis results indicated that the ratio of pumping power over thermal power is less than 1% with an excellent heat transfer coefficient. L. Waganer asked if the separation plate between the hot and cold coolant fluid must be insulated to reduce unwanted heat transfer.

Loss-of-Coolant Accident – Elsayed Mogahed explained how he conducted a loss-of-coolant accident (LOCA) thermal analysis of the ARIES-ST power core. He constructed a FEM axisymmetric (not 3D) model of the core to predict excursion temperatures. He used the 8/25/98 design configuration. He assumed the internal vacuum was maintained for a worst case scenario; hence, heat transfer is through known conduction paths and radiation between elements.

Dave Petti thought E. Mogahed’s groundrules are too severe. He and E. Mogahed will confer on appropriate LOCA groundrules. There was a question on use of integrated heating values, but it was concluded the correct values were being used.

The long conduction path length gives a time constant for the copper centerpost of 93 days. The temperature of the centerpost peaks around 800°C on midplane around 1 day. Mogahad’s Temperature Distribution plots illustrated a sharp discontinuity around one day (the half-life of Cu64 is 12h). Elsayed said that the discontinuity was an artifact of the resolution of the data points selected. He said that the next iteration of results would use a smaller increment in the regions of interest. E. Mogahed suggested the use of a passive heat pipe in the center of the column to help transport the heat to the ends of the centerpost. Dai-Kai said that water would be the best coolant material. The heat pipe approach negates the time delay associated with the long copper conduction path.

Preliminary Safety Assessment of the Divertor – Hesham Khater reviewed his Safety Analysis objectives and assumptions, such as the use of a LOCA accident. He did assume there was still two levels of containment during and after a LOCA. Dave Petti is concerned about the assumptions and will work with Hesham to define an appropriate scenario and equipment status. Dave said that nominal wind conditions may be used rather than worst case conditions.

The team was concerned about the amount of dust sputtered from the divertor surface. ANL is to define the sputtering and vaporization of tungsten during a disruption. The total energy dissipated during the plasma energy quench has been raised to 2 GJ.

Hesham discussed the waste disposal rating data. A majority of the divertor 10CFR61 waste arises from the niobium impurity in tungsten. Still, the divertor plate and manifold will qualify as Class C waste.

Power Core Design Configuration – (no handouts) Xueren Wang reviewed the major elements of the power core and maintenance features. The centerpost should be inverted such that the large centerpost end is at the bottom of the core to allow removal of the centerpost through the bottom. Xueren noted that the radius of the outer PF coils could be reduced to close proximity to the outer TF shell. The three outer PF coils would be structurally integrated, so the only external loads would be gravity loads. The PF coil sizes and locations need to be consistent with the physics recommendations. Both position and shape control coils need to be added inside the TF coils. In a plan view, all TF bus bar connections must be equidistant around the TF shell perimeter for uniform current (and field) distribution. There may be only one NBI injector, but it must fit between bus bar connections and be above the midplane joint. The geometry of the bus bar connections may be altered, but the resistance of each bus bar must be identical.

Poloidal Field Coil Analysis Results – Leslie Bromberg summarized the superconducting PF coil design criteria for NbTi and Nb3Sn, both of which we may use. He discussed his results of the Case J strawman analyses (which differs from the nominal Case 1.6F of 8/25/98 used by most of the design team), noting that the cross-sections for the PF coils are 1/3 of previous calculated values. The innermost coils, PF1 and PF2 are inside the TF coil and need the more expensive Nb3Sn superconductor. Being inside the TF coil shell, they will require dedicated shielding for protection. The remaining PF coils (3, 4, and 5) are located outside the TF coil and utilize an NbTi superconductor. The TF shell is designed to shield and support the outer PF coils. There was a discrepancy between Bromberg’s and Kessel’s calculated current density values arising from the additional helium coolant space in the Bromberg data.

Leslie is still calculating first-of-a-kind costs based upon ITER costs. Ron Miller said he could convert those costs into the appropriate cost basis.

Leslie suggested the TF coil could approximate another PF coil by twisting the TF coil near the ends of the centerpost. Igor Sviatoslavsky said that would be impossible to fabricate, but Les Waganer thought it could be done with his fabrication approach.

Toroidal Field Coil and Centerpost Cost Assessment – Les Waganer reviewed the reasons why the TF coils and centerpost are difficult to fabricate by conventional techniques, which makes them very costly. This high component cost translated into a relatively high cost of electricity. He described two innovative fabrication processes that might help to drastically reduce the cost of the components and the cost of electricity.

The copper centerpost could be fabricated, in place, with a laser forming process that locally melts a metal powder to construct the desired part geometries including coolant channels. The highly automated laser forming process can reproduce detailed parts and good surface finish, but is relatively slow with an estimated build time of 6 months or so for the copper centerpost. Material properties are similar to cast or wrought material. The final unit cost for the centerpost is estimated to be slightly more than $6/kg or $4.4M. The cost of the facilities and hardware to construct the centerpost on site and the cost of energy to run the laser forming process has not been included.

The outer TF coil shell will serve as the return leg of the TF coil and a vacuum vessel. This three-part component can be constructed with a spray cast process, either with aluminum or copper. Aluminum is preferred as it would be lighter on a weight/conductance basis and would be cheaper to build. Les proposed a vacuum-tight inner shell that would be fabricated with a welded half-inch-thick aluminum plate to the final interior shape. The conventional shell construction would cost around $63/kg or $2.5M. This shell will be the preform upon which melted aluminum will be spray cast into place, building up the final shape and thickness with embedded stainless steel coolant tubes. The relatively fast, spray-cast process should complete the three outer TF coil components in less than a month. The spray-cast process unit cost is estimated to be slightly more than $4/kg or $12M.

The total cost for the entire TF coil system is approximately 10% of the cost assuming conventional fabrication techniques. The ARIES team viewed the cost estimates with enthusiasm, but some skepticism. It was recommended that Les determine a few additional cost and process items and data to help "validate" his fabrication process and cost estimate.

Magnet System Design Studies – Wayne Reiersen reviewed several TF coil improvements:

• Continuous shell return leg provides vacuum boundary

• Demountable joint enables power core maintenance from below

• Equidistant bus bar connections on midplane distributes coil current

• Insulating vacuum-tight midplane joint between bus bar connections

• Shell supports all PF coils and protects six outer PF coils

• Flared centerpost or complementary inboard shell collar reduces joule losses

• Tapered centerpost on top is easy to assemble and maintain

• Larger-radius sliding electrical joint located at bottom reduces field and current densities

Wayne showed some details of the tapered upper joint between the centerpost and the TF shell. The geometry used was Case 1.6F (8/25/98), which is a less challenging engineering design. The centerpost radius is increased to 0.9 m and the length is reduced to 17 meters. This reduces the joule heating to 230 MW. This is a good trend, but the final strawman will be different. A CuCrZr material has been chosen for the centerpost material as being a good magnet material and compatible with the new laser forming process.

Wayne presented results of a stress analysis on the Case 1.6F geometry. The stress intensity in the centerpost is low – 53 MPa without packing fraction corrections. The stress intensity in the outer shell is even lower – 23 MPa. Displacements are small.

The outer shell thickness may be increased to help lower the joule heating given that the cost increase is moderate. The PF coil size and location are to be accomplished. The neutral beam penetration needs to be added above the midplane insulator. Additional RF systems must be added for startup and current drive.

Action Item Summaries

Engineering (Mark Tillack)

Design Data Book

- The Design Data Book must be established as the ONLY official data source for ARIES-ST. The Engineering Task Leaders are to input critical system data immediately to assure consistent analyses results for the end of the design phase.


- The divertor substructure will include 13 cm FS and 25 cm high-temperature shield as part of the power core unit. (El-Guebaly, Tillack)

- The 20 cm He/FS centerpost shield is baseline, but analyze two 10 cm D2O cases (35% thermal conversion and zero thermal conversion (throw away power)). (El-Guebaly, Miller)

- Optimize the transition region between the inboard FW and divertor. This includes thermal-hydraulics, maintenance, and shielding issues. In the current design, part of the outboard TF shell in the upper transition region is not adequately shielded to be life-of-plant. (Tillack, Wang, El-Guebaly, Khater, Reiersen)

- Consider new inboard FW coolant routing through centerpost. (Tillack, Reiersen)

- Analyze effect of 3 mm first wall backwall on TBR and inboard shield design. (El-Guebaly)

Blanket and Power Conversion

- Validate core thermal energy balance (inputs versus outputs and losses). (Sze)

- Determine new overall plant thermal efficiency. (Sze)

- Research the compatibility of SiC in hot LiPb. (Sze)

- Incorporate the 3 mm first wall backwall thickness. (Tillack, El-Guebaly)

- Develop the outboard first wall manifold. (Tillack, Wang)

- Examine impacts (on efficiency and tritium) of eliminating He/He heat exchanger. (Sze)

- Incorporate 12 MPa helium coolant pressure. (Sze)

- Reduce pumping power/thermal power to approximately 5%. (Sze)

• Divertor and Vacuum Pumping

- Recommend vacuum path through divertor area and TF shell. (Wang)

- Pump on space behind blanket/shield and inside TF shell. (Wang)

- Determine porous metal heat exchanger divider heat loss and fix, if required. (Tillack)


- Establish credible LOCA and other accident scenarios. (D. Petti)

- Analyze ability of passive heat pipe to transport afterheat during LOCA – Add to design basis. (Mogahed, Sze, Reiersen, Waganer)

- Quantify the radioactive release due to a vacuum boundary failure with INEEL codes. (Petti).

- Establish an atmosphere model (Class F to C). (Petti)

- Calculate the off-site doses under less severe atmospheric model release conditions (Class B or C instead of Class F. (Khater)

- Conduct a disruption analysis. (ANL)

- Use data from the new ANL disruption analysis to finalize the divertor safety assessment. (Khater)


- Locate PF 3,4,5 on outer surface of TF shell. (Kessel, Reiersen, Wang)

- Locate active stability coils. (Kessel, Reiersen, Wang)

- Locate NBI and vacuum duct penetration and shielding requirements. (Mau, El-Guebaly, Reiersen, Wang)

- Determine bus bar locations. (Reiersen, Wang)

- Develop plumbing configuration consistent with maintenance requirements. (Wang)

• PF Magnet System

- Develop full PF coil design. (Kessel, Bromberg, Reiersen, Wang)

- Move PF1 toward midplane. (Kessel, Bromberg, Reiersen, Wang)

Assess capability of twisted centerpost to improve plasma performance. (Kessel, Bromberg, Reiersen, Waganer)

• TF Magnet System

- "Validate" centerpost and TF shell cost estimate (energy, tools, material impurities and control, material properties, recycle). (Waganer, Wille, Reiersen, Miller)

- Assess interface between Al shell and Cu centerpost (corrosion, vacuum welding, disassembly). (Reiersen, Wang)

- Assess fracture mechanics of centerpost and aluminum shell. (Reiersen)

• Heating, Current Drive, and Plasma Startup System

- Develop ECH plasma heating and control system and interfaces with first wall and blanket (Mau, Sze, El-Guebaly, Wang)

Physics (Steve Jardin)

Free-Boundary Equilibrium (Kessel)

- Incorporate plasma squareness constraint and define coil location and current

- Optimize coil locations closer to (or on) TF shell

- Evaluate stability of "best" shape

Vertical Equilibrium (Kessel)

- Finish TSC modifications and evaluate candidate plasmas

- Define appropriate control coils

Startup Systems (Mau)

- Model Politzer scheme with TSC code

- Define appropriate ECH system

Current Drive Systems (Mau)

- Assess diamagnetic plasma flow

- Do a full calculation of NBI CD system including size of beam and intercepting angle

Divertor Region (Petrie, Jardin)

- Determine distribution of transport power (80% outboard, 20% inboard uniformly over full length of centerpost?) and radiation power

Plasma Disruption (Jardin, ANL)

- Determine plasma disruption conditions - total stored energy and frequency (one per year?)

ARIES-ST Project Meeting 13 October, 1998

Thursday, 17 September 1998

Princeton Plasma Physics Laboratory


8:00 Coffee and Socializing PPPL hosts

8:30 Administrative Details Jardin

8:35 Welcome to Princeton Goldston

8:45 Review of Agenda Waganer

8:50 Project Status and Plans Najmabadi

9:05 DOE Update Dove

Physics Session

9:20 NBI's Roles in Current Drive, Rotation Generation and Startup Mau

9:40 GA Physics Studies Results P.Politzer

Including bootstrap overdrive startup, transport, divertor physics, equilibrium, and stability

10:25 Break All

10:40 The Dependence of Beta on Aspect Ratio for STs Menard

10:55 Equilibrium and Vertical Stability Studies Kessel

11:15 Time-scales for Current Ramp-up in STs and Tokamaks Jardin

11:35 Economics of Alpha Channeling in Tokamak Power Plants Ehst

11:55 Discussion of Physics Design Point Led by Jardin

12:30 Lunch

Systems Studies

1:30 ARIES-ST Systems Study Results Ron Miller

- Discussion of Strawman Cases

- Impact of Design Choices and Operating Regimes

- Recommendation for Final Strawman

2:15 Discussion of Final Strawman Led by Miller

Engineering Session

2:30 Inboard Shielding Options and Impact on Design El-Guebaly

3:15 Inboard Shielding Design Sviatoslavsky

3:30 Break

3:45 Blanket and Power Conversion System Summary Sze

4:15 Divertor Design and Blanket Analysis Tillack

4:45 Preliminary Safety Assessment of the Divertor Khater

5:15 Extra General Discussion Time All

5:30 Adjourn


ARIES-ST Project Meeting 13 October, 1998

Friday, 18 September 1998

Princeton Plasma Physics Laboratory



8:00 Coffee and Socializing PPPL hosts

8:30 LOCA Analyses Using an Axi-Symmetric FEM Mogahed

9:00 Power Core Design Configuration Wang

9:30 Results of PF Coil Analyses Bromberg

9:45 Break

10:00 TF and CP Cost Assessment Waganer

10:45 TF and CP Design Approach Reiersen/Dahlgren

11:30 Extra Discussion Period (15 min)

11:45 Lunch

12:45 Wrap-up of the Design Approach for ARIES-ST Tillack

1:05 Wrap-up of the Physics Approach for ARIES-ST All

1:25 Next Actions to Prepare for Design Completion/Next Meeting? Najmabadi

1:50 Break


Overview of ARIES-ST Power Plant Design

2:00 Summary of ARIES-ST Power Plant Najmabadi

2:20 Summary of Physics Parameters and Systems Jardin

2:40 Summary of Engineering Design and Configuration Tillack

3:00 Adjourn