ARIES Documents -- Meetings ArchiveARIES Project Meeting, 18-20 September 2000
Documented by L. Waganer
Encl: Action Item List
AdministrativeBill Dove informed the team that the 2001 fusion budget has not been approved through conference committee yet. If necessary, the smaller Senate budget version should be used for planning purposes. Bill noted that he had received most of the ARIES Peer Reviewer comments and will soon assemble a summary report of the final comments and recommendations. No details can be released at this time but the reviewers were very complimentary. One general comment was that the ARIES team should be more proactive on distributing its results and findings to the fusion community in general. Farrokh Najmabadi also affirmed the general positive comments with good suggestions for improved communication.
Farrokh Najmabadi highlighted the need to complete the ARIES technical work as soon as possible and commence the final report chapters to be published in the FED journal. A tentative date of 1 January 2001 was adopted. The ARIES-ST Final Report chapters are all complete and in the hands of the reviewers (except for the systems study results). The recommendations from the reviewers will be returned by 15 November.
Steve Jardin will work with his physics group to recommend areas of MFE physics concentration for next year.
ARIES-AT Final Design Presentations
Physics AnalysisOutline of Physics Final Report Chapter -- Steve Jardin presented a proposed outline of a physics chapter for the FED journal. This chapter would highlight the physics goals and results for the ARIES-AT design. Steve compared the ARIES-AT goals to the prior ARIES-RS study. In general the AT goals were more aggressive and the analyses were more complete and rigorous. Steve described remaining areas that could be analyzed, improved, and verified with testing. The main chapters suggested are:
Steve noted that new information and data are being offered by GA concerning applicability of H-mode plasma edge physics on the ARIES-AT in place of the current L-mode assessment. It was suggested that the final report contain a discussion of how an L-mode plasma would impact current drive capability, plasma confinement, and divertor design and operation.
There will also be a discussion in the chapter of stabilizing resistive tearing modes (RTM), either by plasma rotation or active stabilization coils. Neoclassical tearing modes (NTM) will be discussed along with island formation.
The recommended current drive systems for ARIES-AT will be presented. The ICRF-Fast Wave (FW) is the system of choice for on-axis current drive. Lower hybrid wave is chosen for off axis current drive (CD). It is also recommended that a high harmonic fast wave be retained as an alternate off axis current drive system.
The ARIES-AT plasma transport and energy confinement analysis will be compared with the community effort to calculate from first principles. The first principle results compare to experimental results to within 20%.
The present ARIES-AT design does not require plasma rotation to stabilize NTM. The current analysis of the ARIES-AT edge physics indicate that approximately 30% of the transport power is radiated to the wall and the remainder in the radiative divertor. Injection of neon, argon or krypton would be used to control the desired radiation level.
It was recommended that a section be added to the physics chapter to help guide the fusion community on necessary research and development to demonstrate or validate the ARIES-AT physics basis.
Evolution of AT Physics from ARIES-RS to ARIES-AT -- Chuck Kessel highlighted the AT plasma improvement and changes from ARIES-RS, especially in the area of plasma equilibrium and free boundary conditions.
The 99% free boundary flux surface was used to improve the fidelity of the modeling and predicts higher beta from the stronger shape parameters. The more flexible pressure profile (modeled with extra terms) allows simultaneous beta and bootstrap alignment optimization.
Increased triangularity increases betaN and plasma current, Ip. Previously the neutron damage to the inboard superconducting TF magnet from the radiating divertor slot limited the triangularity. Experiments and simulations show that no slot is necessary to obtain detachment on the inboard side, thus allowing increased triangularity. The AT plasma also has increased elongation that also increases betaN and Ip. This increase in elongation was made possible by moving the tungsten stabilizer between blanket elements and closer to the plasma edge. The current drive system does not need a HFFW subsystem because the bootstrap current is better aligned from the extended and improved pressure profile.
The lessons learned from ARIES-AT physics assessment are:
Heating and Current Drive Systems for ARIES-AT -- TK Mau told the group that the latest series of ARIES-AT equilibria have profiles optimized for high betaN (90% of beta limit) and maximum bootstrap alignment (Ibs/Ip > 0.9) at Zeff = 1.7, Te0 = 24, 26, 28 and 30 keV. The bootstrap current is sensitive to Zeff. TK extrapolated from the nominal Zeff of 1.7 by adjusting both the plasma density and temperature profiles. He was able to determine good alignment of the bootstrap current and only needed two current drive systems for the current profile. At a Zeff of 1.8, three CD systems would be required. TK explained the design of the ICRF fast wave and the lower hybrid wave launcher subsystems. Both could be located in a single power core sector.
Summary of Next Year's MFE Physics Activities -
Les presented the results of the assessment of the maintenance approaches that concluded the best approach was the refurbished sector from the hot cell because of faster replacement times, reduced contamination of the maintenance areas, and higher quality of refurbished components. This approach has a disadvantage of higher waste volume involving extra blanket and shielding structures. Additionally, a cask approach was adopted to transport the vacuum vessel and power core sector to the hot cell.
Les presented an assessment of the frequency of power core maintenance actions, with the selected choice as replacement of one half the power core every two years. Based on the prior analyses, Les presented the development of the power core building configuration and the sequence of activities to remove and replace the power core.
He also showed a timeline of the times to shut down the power core, start up the power core, and to conclude the repetitive actions to remove and replace a single sector with a single cask and transporter. A trade study was conducted by adding transporters and sectors to determine the optimum number of maintenance sets.
Availability goals were assigned to the major plant elements to determine if the power core could achieve the necessary plant availability goal of 90%. Les postulated that it is possible to achieve the 90% value. The ARIES team recommended that the power core unscheduled annual outage goal of 10.28 days/year be doubled to 20.56 days/year. This would result in a plant availability of 87.6%, which was adopted for the ARIES-AT baseline. If any relevant maintenance or availability data from the APR becomes available, the availability number would be reanalyzed.
Final TF and PF Coil Design and Analysis Results -- Tom Brown displayed the most current CAD modeling of the TF and PF coil designs. He noted where the coil design had been modified to address the stress and deflection findings from Fred Dahlgren. The main problems arose at the upper and lower inner corners of the TF coil. A slightly larger radius would lower the stress levels.
Final TF and PF Structural Analysis Results -- Fred Dahlgren expanded on the TF stress and deflection analysis findings. A slight modification in the structural configuration would bring the deflections within design allowables. He also analyzed the worst case field perpendicular to the high temperature superconductors. The PF coils 12, 13, and 14 are at the limit of the permissible fields.
Final HTS Coil Definition -- Leslie Bromberg reiterated that the high temperature superconductor (HTS) is generally comparable to the low temperature superconductor. The HTS does have an advantage that no coolant is required internal to the superconductor pack. The thermal capacity of the HTS is so large that the quench protection only requires thin current switches. Leslie predicted the fabricated cost of the HTS coils would be approximately $50/kg.
Waste Issues and Radiological Inventory in LiPb -- Laila El-Guebaly noted that the Nb alloying element in the Inconel structure of the coils is a waste problem as it contains thirty times the concentration of Nb as does 316 LN-SS. But this is a correctable problem, which can be eliminated by altering the composition of Inconel or choosing a different structural material. She noted that there are safety concerns regarding both Po and Hg contained in the LiPb coolant inventory. Lead generates Bi, Po, and Hg. Polonium can be controlled either by limiting the concentration of bismuth in the coolant (which produces Po) or removing Po directly; whichever is the simpler to do. Laila explained different calculation techniques to estimate the bismuth and polonium inventories as a function of the time in the reactor. She concluded the LiPb coolant should be processed to remove Po and Hg whenever the plant is operational to maintain acceptable concentrations. Dai-Kai Sze requested the maximum generation rate of Hg to determine processing capabilities.
Heat Transfer System and Coolant Processing System -- Dai-Kai Sze discussed his results to control the production and concentration of Po within the LiPb coolant system. Po is produced mainly from Bi209, which in turn is produced from Pb. The concentration of Po can be controlled either by controlling the concentration of Bi209 or directly controlling the concentration of Po. Controlling Po is much easier to do. Dave Petti suggested controlling the concentration of Po to 0.001 wppb. Dai-Kai explained that experimental evidence supports that a diffusion process will be sufficient. The extraction of Po can be accomplished in the tritium recovery system.
Dai-Kai is beginning to assemble information on the Hg cleanup system. The end-of-life inventory is 88 g, but the value of the maximum production rate (15 microgram/s) is needed to determine the system capability.
Final Safety Studies Results -- Dave Petti reiterated the safety limit of less than 1 rem release so that no public evacuation plan is required. He noted several accident scenarios, which have been investigated, produced no adverse safety concerns. The current accident scenario being investigated is one of a water leak from the vacuum vessel onto the backside of the high temperature shield containing LiPb.
Plans for the ARIES-AT Final Report -- Farrokh Najmabadi informed the group that the recommended report format can be found on the ARIES web site. He would like a complete title and the first three authors of the ARIES-AT papers to be published in the FED journal ASAP. The figures should be postscript, set at 1200 dpi (not 300 dpi). Use color sparingly as it costs quite a bit to print a page in color.
ARIES-IFE Study Presentations
IFE DriversHeavy Ion Drivers -- Ed Lee informed the group that a lot of the heavy ion fusion (HIF) work that applies to commercial power plants is being done at the HIF Virtual National Laboratory (VNL), which is comprised of LBL, LLNL, and PPPL. An HIF VNL web site is being constructed to help disseminate the results to the fusion community. Roger Bangerter is the temporary VNL director.
Ed discussed some of the technical details of the proposed target options and the direction and extent of the research. He illustrated the trend in target physics by comparing different types of indirect drive heavy ion targets and their respective yields.
Ed discussed in more detail the driver requirements for the target (energy, ion range, ion species, charge state, possible ion sources, and expected driver efficiencies). He examined the most promising driver beam arrangement. He stressed the importance of obtaining a space charge potential much less than the beam temperature. The final focusing magnets are well protected because they are located roughly 7 meters from the chamber wall.
Ed discussed the near term experiments and the overall HIF project strategy leading to a demo plant. The Integrated Research Experiment is one of the key experimental facilities being proposed. It is a high current (10 to 100 A) transport experiment to develop the key driver technologies.
Heavy Ion Driver Systems Modeling -- Wayne Meier informed the group that he has continued to develop and update a heavy ion (HI) driver systems model with current improvements in transverse and longitudinal emittance growth and models for the final focus, quadrupoles, and drive compression. This HI driver model could be readily applied to this study with ARIES efforts focusing on costing, energy storage, pulsed power, and cryogenic systems. Existing models can be updated to handle the target gain scaling, chamber models, power conversion, and BOP systems.
Wayne then presented his integrated systems analysis results for a 3.3 MJ,
Rb+1 driver design that had been presented at the HIF symposium in
March 2000. Areas with high payoff were judged to be target improvements, high
acceleration gradients, and core performance. High magnetic field gradients (2
MV/m) have resulted in shorter driver designs (< 1 km) than past designs.
Using this basic driver design, he presented several parametric variation in
the key driver parameters, such as ion energy, lattice half period, core inner
radius, and number of beams. He found that the inner core radius is minimized
with a quadrupole field in the range of 4 to 5 Tesla. He also found that a
minimum of 160 beams is needed to meet the spot size requirement. The direct
capital cost of such a driver system would be in the range of $0.7M, with a
minor variation (10%) for (single) design point variations of 30%.
Output Threat Spectra from Direct and Indirect Drive IFE Targets -- John Perkins compared the design and constituents of the DD and ID target capsules. Output energy levels for both targets were catalogued by X-rays, neutrons, gammas, fast ions, debris ion kinetic energy, and residual thermal energy. The X-ray, fast ion, and debris ion kinetic energy spectra were given over a range of respective energy levels.
Final Transport of HI beams -- Craig Olson informed the group about the necessary conditions for HI beam transport through the chamber environment to the target. The necessary HI beam perveance is 1.6 x 10-5 (dimensionless?). With a hard vacuum, ballistic transport of 500 beams is possible with a perveance of 1.6 x 10-6. Ballistic focus can also be done in atmospheres of 10-4 torr to 10-3 torr with perveance in the range of 1.6 x 10-4. For ballistic focusing, the charged beam used in the accelerating modules must be neutralized. There are several techniques of neutralizing the beam. In the region approaching 1 torr, the beam will strip and the more likely transport is the self-pinch mode. Craig summarized different beam transport techniques and chamber pressures assumed in prior studies.
Plasma Channel-Based Final Transport -- Simon Yu informed the group that the channel transport technique could match well with dry wall chamber concepts since they might have some pressure in the chamber. It is also insensitive to chamber size as opposed to ballistic, which is restricted to smaller radii chambers. The smaller size of the chamber penetration reduces the neutron streaming through the penetration. However, there is more risk associated with a plasma channel approach. There are two classes of plasma channel concepts: preformed channels and self- pinched channels. Several lasers create a conductive channel that guides the beam to the target. In addition to the single channel to the target, there also has to be a channel for the return current, so additional channels are necessary at right angles to the first channel, intersecting at the target. Channel transport has successfully produced a 55 kA channel with a 4 mm radius. Z-pinch channels are possible and stable z-pinch channels have been produced at 4 mm diameter, 40 cm long, with 60 kA
Simon presented preformed-channel transport design parameters to yield prepulse and main pulse Pb beams for the chamber conditions consistent with HYLIFE-II with 5 torr Xe. He also showed the reactor and final focus schematic and the wall lens and insulator shields.
Simon noted that the optimization of beam transport is in work, including self-fields, transport efficiency, spot size, and sensitivity to electron temperature.
Output Calculations for Laser Fusion Targets -- Bob Peterson explained the variables to be considered in choosing the gas environment in a chamber. Bob used the BUCKY 1-D radiation-hydrodynamic code to simulate the target, gas behavior, and wall response. He uses this code with both direct and indirect targets with both laser and HI beams to determine target output. He contrasted the two different SOMBRERO (1991 and 2000) targets with the newest NRL targets. He noted the SOMBRERO DD target output is dominated by neutrons and energetic ablator ions. The BUCKY code does not have zooming, so laser deposition does not exactly agree with codes that model zooming. So Bob adjusts the contrast between foot and main pulse to correct the data. He showed several time-varying parameters after implosion and ignition. He presented ion spectrum from the BUCKY results. There seemed to be some correlation with the LLNL code results presented by John Perkins, but the gold ions were in error due to the code constraints (probably not as energetic as shown).
Parametric Studies of Dry-Wall IFE Chamber Dynamics: Xe Pressure and Chamber
Radius -- Bob Peterson showed the results of parametric surveys of dry wall
chamber dynamics (blast wave propagation and first wall vaporization) for both
SOMBRERO and NRL targets. He found that the first wall vaporization depends
critically on both the xenon pressure and chamber radius, as it is necessary to
keep prompt x-ray fluence below a critical threshold value. Propagation of the
blast wave depends on the opacity of the chamber gas. For direct drive targets
in a SOMBRERO type chamber, radiation flow is governed by emission, not
transport. At 0.5 torr Xe chamber gas conditions, the larger yield SOMBRERO
targets launch stronger shock waves through the gases than does the NRL target.
Bob showed the vaporized wall masses as a function of the chamber gas density.
The threshold was below 0.5 torr for the SOMBRERO target and 0.15 torr for the
NRL target. At 0.5 torr, the chamber radius was varied and the results
indicated that a minimum of 4.5 meters would be the threshold radius.
Shielding considerations for HIB with Preformed Channels -- Laila
El-Guebaly presented Sawan's shielding analysis for the dry wall chamber of the
HIB driver with pre-formed channels. A one-meter thick dry wall chamber
provides a lifetime protection for the insulators. The final focus magnets
were analyzed to determine neutronic effects. An additional 35 cm thick local
shield would be needed to protect the FF magnets. It was suggested the laser
final mirror be moved to 25 meters. Also the next turning optic should be
The warm assembly process has the advantage of easing the complexity of the hohlraum assembly equipment but has the distinct disadvantage of significantly increasing the tritium inventory required for the target fill facility. Unless methods can be found to reduce the tritium inventory requirements for permeation filling, this may dictate selecting the cold assembly process. Either process could use the "temperature shimmed hohlraum" (TSH) technique, developed for NIF cryotarget fielding, for layering in the hohlraum. A task to evaluate thermal requirements for high-volume TSH layering and provide design concepts for a mass-production system is starting. For the direct drive radiation preheated target design, key fabrication technologies have already been identified and work is beginning to enhance the process toward the commercial scale. An experimental program to evaluate fluidized beds as a potential technology for capsule production is also getting underway.
IFE Target Activities -- Pete Gobby listed the critical issues to consider in the target fabrication process: materials, processes, cryogenics, shielding, capital costs, and operating costs. Pete examined the fill time requirements for typical capsules using a just-in-time process control approach. For GDP capsules, the minimum fill time is around two hours. The buckle strength and the permeation for the capsules establish the fill time due to delta pressure constraints. The NRL capsule wall is much thinner, hence the delta pressure is smaller and the fill time might be as long as five days.
Target Injection in a Gas-Filled Chamber -- Dan Goodin said that GA and UCSD have been working on measuring the reflectivity of thin gold layers fabricated with current equipment, as a function of layer thickness, incident radiation wavelength, and angle of incidence. Current target heating calculations have assumed approximately 98% reflectance of the blackbody radiation, admittedly a high value to achieve in routine production.
For studies of target heating during injection, GA is using the NRL target as the baseline design with a pressure in the chamber ranging from vacuum up to the Sombrero reference value of 0.5 torr. Target injection speeds up to about 400 m/s are being considered. The current strategy for dry-wall chambers is to try to reduce the gas-pressure to reduce the extent of total target heating and to avoid the asymmetric convective heating during injection. It appears that reduction of the gas pressure to about 5 mtorr may be necessary to control target heating with the current thin-wall target designs.
The traditional concept for tracking targets is to track them outside the chamber and predict their trajectory to target chamber center. But the gas environment inside the dry-wall chamber may influence the trajectory in a non-predicable way. An initial assessment found that at 50-mtorr gas pressure, the variability of the gas density from shot to shot must be less than 0.01%. Alternatively, the absolute gas density must be measured to this accuracy. Given this sensitivity of the target trajectory to gas variations, they are looking at methods to track the target within the chamber to near the final location.
Design of the Target Injection and Tracking Experimental System -- Ron Petzoldt described the strategy to develop an experimental target injection and tracking system. The critical issues are:
Various injection methodologies have been evaluated and a light gas-gun was selected for the experiment, along with 1-D photo-diode array detectors for tracking. Ron stated that a Conceptual Design Review for the equipment would be held in San Diego on September 27, 2000. Ron stated the goals for this facility were to:
Ron said GA was following a program approach of:
Don intends to compile a complete list of system conditions that might result in a significant off-normal event. He would then look at both the probability of the "failure" and consequences of those events.
IFE Systems Activities -- Ron Miller discussed the adaptation of elements of the ARIES Systems Code to the IFE assessment study. Wayne Meier's driver model will also be incorporated along with capsule and chamber models. Ron discussed several commercial software packages capable of working with risk and uncertainty models.
Target Gain Curve Modeling -- Wayne Meier presented information on target gain versus driver energy curves for direct drive target with lasers and indirect drive with heavy ion (HI) drivers. The laser gain curves are simple fits to published gain curves. They cover a rather broad range depending on success in achieving a low-adiabat implosion. The highest laser gain curve goes through the NRL point design of G = 135 at Edriver = 1.2 MJ. The spot size and focusing requirements will be determined as a function of driver energy for the laser gain curves for input to the driver focusing and target injection system requirements.
Curves for distributed radiator targets were also shown. The results are base
on scaling relations given by Debbie Callahan. The key design variable for the
heavy ion targets (beside driver energy) is the ratio of the hohlraum size to
fuel capsule size. Close-coupled targets use smaller hohlraums for a given size
capsule, thus less driver energy is needed to reach the required drive
temperature. As a result the close-coupled target gives higher gain, but they
require focusing to smaller spot sizes. These preliminary results will be
reviewed with Callahan-Miller before using them in the systems modeling.
The next concern arises from the iodine and cesium isotopes contained in the xenon chamber gas. A chamber gas cleanup system can remove those isotopes and reduce the dose. Alternatively, a lower activation gas, such as krypton, could be used as the chamber gas.
Safety Activities (Petti)
Nuclear Analyses (El-Guebaly)
Target Injection and Tracking (Goodin)
Target Fabrication (Goodin)
Chamber Engineering (Raffray)
Final Optics Assessment (Tillack)
HI Driver (Yu)
Target Chamber Analysis (Peterson)
Safety Activities (Petti)
Nuclear Analyses (El-Guebaly)
Target Injection and Tracking (Goodin)
Target Fabrication (Goodin)
Chamber Engineering (Raffray)
Final Optics Assessment (Tillack)
HI Driver (Yu)
Target Chamber Analysis (Peterson)