Guests: V. Chan (GA), E. Cheng (TSI Research), R. Doerner (UCSD), R. Fouch (UW), J. Leuer (GA), M. Peng (ORNL), A. Turnbull (GA), A. Ying (UCLA)
Martin Peng explained his design basis for the LAR tokamak fusion power core. Much of the LAR physics are based upon the successful Culham START experiment. This experiment illustrated that the magnetic field lines become longer on the inboard (stable) region as the aspect ratio is lowered. Additional coils can shape the plasma from an elongated shape to a triangular shape. He presented a chart suggesting that LAR plasmas would have no disruptions, but the audience recommended this statement be qualified to the START experimental results or to certain qedge or qon-axis conditions. M. Peng thought the START results were encouraging but Bob Conn cautioned that the data might be misinterpreted due to the relatively cold plasma. The design basis and expected performance for the bare centerpost was discussed. The water-cooled copper temperature was 150 C at the midplane. Results from the SUPERCODE were shown comparing several plasma concepts including ITER-like, TPX, and an enhanced reverse shear plasma. It was suggested that additional information, such as recirculating power data and current drive requirements, would be helpful in deducing conclusions and meaningful trends.
Steve Jardin discussed the results of the assessment of several LAR plasma concepts. The areas of investigation were plasma stability and current drive requirements over an extensive parameter space. He discussed a series of plasma configurations as to their stability, beta capability, and current drive requirements (on-axis and off-axis). One particular configuration (A=1.25 and k=2.8) was very attractive as it was a stable configuration with a beta of 35% and practically no current drive required with a bootstrap fraction of 99.5%. F. Najmabadi asked if this plasma configuration was operated at 95% of theoretical beta, would the profile and CD requirements significantly change?
Chuck Kessel explained his analyses to determine plasma stability and vertical stability. He assumed all derivatives of dB/dR should be ~ 0 which can be examined using a pure vertical field to examine the plasma shapes. His results were incorporated into the results shown by S. Jardin.
T.K. Mau presented his RF current drive (CD) assessment for LAR power plants. He pointed out the unique features of LAR that limit the RF technique to low frequency fast waves. Current drive off-axis and near the edge is most inefficient and should be avoided. As a result, equilibria with good current profile alignment in the outer plasma region are preferred. In that case, CD power in the 10's of MW is possible.
Vincent Chan of GA informed the group about his spreadsheet analyses of the LAR physics. He felt his results were consistent with the ORNL results of M. Peng. V. Chan outlined an alternative development path which envisioned the use of the LAR concept on a high risk, high payoff route from a Proof-of-Principle experiment to Pilot Plant to Demo. He felt that high normalized beta is the key to simultaneously achieve high fusion power and high bootstrap current fraction. He assumed an algorithm fit of beta normal = 12/Aspect Ratio. M. Peng took exception to that algorithm fit and suggested a coefficient of 10 instead. From the engineering side, V. Chan assumed a 7 meter copper centerpost with a life of two years. The LAR is neutron wall loading (NWL) limited to a value of 8 MW/m2. His resistive power losses in the TF magnet central column were in the range of 40 to 85 MW for a pilot plant which was felt to be too low. Some of the audience also felt the NWL of 8 MW/m2 was too high. Clement Wong indicated that he already has a design that can handle this high NWL. Jardin, Chan, and Peng agreed to cooperate to resolve physics issues. F. Najmabadi also offered the limited support of the Starlite Engineering Group to investigate specific issues.
Mark Tillack reviewed the Starlite Team's results of examination of certain engineering issues, such as the properties and performance of the centerpost and the high power handling capabilities of the blanket and divertor. Even a thinly shielded (30 cm) center post seems to exhibit a much improved lifetime, lower waste concentrations, and better resistivity values. The estimates of the recirculating power values were at a much higher level than those predicted by Peng and Chan, mainly due to the power dissipation in the TF coils and leads. It was also felt by the Starlite team that the power conversion needs to be in the range of 46% thermal conversion efficiency and this figure can only be reached by using liquid lithium or helium coolants.
Chuck Bathke reviewed the structure of his systems code, the major inputs, and relevant results. He stressed this was not a point design, rather was a parametric trade study with key physics and engineering inputs and constraints. The plasma cases analyzed by S. Jardin et al. are compared in the code with appropriate values of beta and current drive. The centerpost has a 30 cm inboard shield in order to meet the Class C waste requirements. The blanket is assumed to be a liquid lithium breeder and coolant in a vanadium structure. The center post properties were input or derived. Trade studies were conducted over a range of aspect ratios, minor plasma radii, plasma ion temperatures, TF coil properties, a current drive efficiency, and center post shielding effectiveness.
General - Bob Conn reviewed the findings of the recent FEAC deliberations. He explained how the FEAC considered the inputs received from all the constituents. All voices were heard and considered. He discussed how the Starlite team might contribute in the coming years to help guide fusion in the era of Fusion Energy Science.
Bill Dove discussed his understanding of what future budgets and the program direction might be. He also alerted the project of a potential budget reduction of $125K for FY96. He stressed that we should try to inform the DOE program office of our results by holding a project meeting in the Washington area. F. Najmabadi also will investigate giving a project briefing to DOE leaders in the near term to help them understand our project goals and results to date.
High-Level Status of the Reverse Shear Design Concept
Chuck Bathke presented the geometry results from the recently completed RS strawman. The bucking cylinder has been extended to form a cap and the upper PF coils have been moved to accommodate the cap. The inboard PF coil stack has been modified to be continuous. C. Bathke summarized the design rules that were integrated into the code to define the first wall surfaces in the divertor region. C. Bathke also mentioned that the a-pressure modeling of the current drive system efficiency is changing, which will increase the cost of Second Stability plasmas, e.g. ARIES-II. Chuck also has improved the fidelity of the blanket and shield volume calculation routine. He illustrated the change in COE due to the location of the stabilizing coil (b/a). This COE effect will change due to the integrated effect of the vertical and kink shells.
Steve Jardin summarized the RS plasma configuration assumed for the nominal case, which has a kink stabilizing shell located at 1.3. A study was presented showing that plasma triangularity has a strong impact on MHD stability for reversed shear plasmas, and that for triangularities below 0.4, the normalized beta drops rapidly. After discussion it was decided that the physics group will redo the baseline with a lower elongation (kappa at the x-point about 1.9) to allow the vertical stabilizing wall to move back behind the reflector. This new baseline will benefit from having the kink-wall moved in to 1.1. Steve also summarized a list of critical physics issues which were later prioritized.
Dave Ehst presented the results of his analyses of the stability and current drive on advance plasma concepts. His interest in this area arose when the economics of reverse shear and second stability power plants seemed to be similar, but the plasma differences suggested that reverse shear should be superior. The equilibrium analyses conducted for the second stability ARIES-II used the then- best available technique, namely the Harris model. Now, the full Hirshman-Sigmar model would predict a different result which would degrade the bootstrap current and impact the required current drive and raise the COE. The project was concerned that we should use the best analyses tools and not report out-of-date results that would infer incorrect conclusions. On the other hand, rerunning old cases with new tools would not yield correct results as the re-examined cases should be completely reoptimized. The only proposed solution that minimized the effort involved was to CLEARLY and EXPLICITLY footnote any comparative results.
Mark Tillack discussed the status and future tasks of the engineering group. The principal activities fall into three categories; (1) analysis, (2) addition of design detail, and (3) integration. The final design is expected in July. Therefore, we should be at a level of change in the design of no more than a few percent by the next meeting. Some of the key design features which must be resolved immediately are the divertor pumping ducts and shielding and the maintenance port layout.
[Moved some Friday talks into this time slot.]
Don Steiner mentioned that the shuffling of budgets and personnel has been completed. Fred Silady of GA, Greg Hofer of Raytheon, and Robert Thayer of RPI will work on the preliminary safety analysis. The initial assessment will be based upon the similar ARIES-II power core to try out the process. Results of the ARIES-II analyses will be presented in the March meeting for comment. If favorable, the design basis will be modified to the new RS configuration with results to be reported in the Starlite final report (Fall 96).
Ron Miller reviewed the status of the planned Starlite economic modeling. He also mentioned the benchmarking of the code inputs and outputs with available utility data and other economic models. He was encouraged to pursue exchange of data, models, and results with Ian Cooke of UKAEA.
Laila El-Guebaly informed the project of Jake Blanchard’s effort to obtain and apply codes to model electomagnetic (EM) forces on representative blanket structures. Jake has obtained the CARIDDI code to model currents and loads and also has the commercial ANSYS code that will also determine the resultant stresses. The ANSYS code is a more universal FEM code that may be easier to use. S. Jardin asked why the PPPL SPARK code was not considered. He suggested cognizant PPPL personnel could be funded to conduct the analysis, but the cost to the project was a concern. Jake Blanchard will follow up on this suggestion.
Discussion/Resolution of Significant Issues
Implementing a Conducting Shell [Led by D-K Sze]
Dai-Kai Sze reviewed the general need for a conducting shell and/or coils to be located within the blanket and the engineering constraints that limit their application, such as placement, cooling, materials, effective conductance, power, and waste issues.
Chuck Kessel discussed the need to have a conducting shell to provide kink stability. The shell need only be located on the outboard wall (Ī 80° from midplane) and may have toroidal breaks such as 1/16 modules. It was agreed to evaluate the kink stability conducting shell as the second layer of vanadium (2 cm thick and 3.5 cm from the first wall surface). This would roughly correspond to a b/a of 1.1 but D-K Sze will determine the value to be used by C. Kessel to re- evaluate the effectiveness of the vanadium kink stability shell close to the first wall.
A second conducting shell is needed for vertical stability and an active control coil, both located within the blanket, preferably around a b/a of 1.45. The shell and coil are to be either continuous or act as such (saddle coils). The engineers preferred the shell to be behind the reflector, which is at a b/a Ň 1.5. The engineering preference for the coil location would be outside the shield. The suggested compromise is to use a 3-cm-thick tungsten shell, operating at Ň 650°C behind the reflector at a b/a = 1.5. The coil will be located outside the cold shield, will be either continuous or a saddle coil. The triangular coil suggested by Igor Sviatoslavsky will not work. Chuck Bathke should use 10 MW as a nominal power consumption, but C. Kessel will try to verify. D. Steiner will check on the lithium-based IBC coil for applicability.
Dennis Lee presented Igor Sviatoslavsky's material on stabilizing conducting shells and active coils. Some of the material was not germane because of the decisions on the conducting shell, but some was still relevant. Igor had suggested a triangular-shaped coil but the return legs would tend to cancel the coils at the surface. Thus the return legs need to be located further out or connected to the adjacent module coil. The foil approach for connecting adjacent conducting shells was not thought to provide adequate conductance and might weld closed.
Divertor Design [Led by C. Wong]
Clement Wong summarized the design status of the divertor system. The design presented represents his integration of the design, based on a radiative divertor approach. The configuration selected was a natural open divertor based on the contour of the equilibrium field lines and the necessary opening to provide minimum material erosion. The depth of the outboard channel of about 1 m is to help trap neutrals and reduce electron temperature and sputtering at the target plate by allowing significant radial energy transport into the private flux region below the X-point. The deep divertor channel provides good isolation between the divertor and the core to help keep impurities entrained and it provides necessary distance for momentum loss via charge exchange. Adequate pumping should be designed to help trap the injected impurity gas in the divertor region. The inboard channel has a shallower channel for power handling, which is possible because of larger magnetic flux expansion and the reduced power flow to the inner leg in double-null plasma.
For the analysis of the design, Clement began by stating the incomplete understanding of the divertor design. No simple accurate modeling approach is available and sophisticated modeling tools are only as good as the inputs for the codes which are not ready to be used as predictive tools. To estimate the power flow and the maximum divertor heat flux, the approximation was based on the operating experience of a double null experiments’ database from ASDEX, PDX and DIII-D. This was the same approach used in the scoping for the PCAST divertor design. With the assumption of a 1 cm mid-plane power flow SOL thickness and different radiation fractions in the core to the outboard and inboard SOL and at the divertor, the maximum heat flux to the divertor was calculated. Excluding radiation effects from additional impurities and taking into consideration a triangular distribution of heat flux at the end of the separatrix, the maximum divertor heat flux can be as high as 50 MW/m2, which is an order of magnitude higher than the engineering goal of 5 MW/m2. To estimate the radiation effects from the addition of impurities, separate effects from the mantle and the divertor will be evaluated. The physics modeling of this approach is continuing. Due to the necessary impacts on SOL temperature and density, which in turn will impact the core energy balance, bootstrap current and current drive parameters, Farrokh suggested preliminary inputs be provided in two weeks in order to initiate the next round of a system code study in a timely manner. Clement will take the responsibility of fulfilling this project need.
A coolant routing design was selected. Preliminary heat transfer results indicate the possibility of meeting various materials design limits with a coolant outlet temperature of 450 C. A higher coolant outlet temperature will be evaluated to enhance the thermal efficiency of the design. A 2-mm layer of tungsten was suggested on the divertor high heat flux regions to minimize the surface erosion rate. The concern about the release of radioactivities during a major accident condition was raised. The inclination of the target plates was selected to be 15-deg to reduce the local heat flux. Based on the strawman results, the pumping speed at the divertor throat was estimated to be 90 m3/s. Clement stated that we will need supports in the area of vacuum system design and the structural design of the divertor. We will first evaluate the radiative approach; additional physics assumptions, e.g. neutral gas injection and detached plasma, may be needed to come up with a robust divertor design.
Dave Ehst discussed the ELM disruptions and how they affect the first wall surfaces. They are disturbances that could occur for the reverse shear mode, but they are less severe (1 to 6%) than a major disruption such as a vertical displacement event (VDE). There are several varieties, such as Type I, Type III, mossy, large, and grassy. The various types are distinguished by the frequency and power flow. Dave recommended that, if ELMs are unavoidable, the best type to encourage would be the grassy ELMs which have a high frequency (≥ 1000 Hz) and a low power discharge per pulse.
Power Core Configuration [Led by D. Lee]
Dennis Lee started the discussion off by summarizing the main issues to be addressed in adopting a baseline configuration. C. Bathke reviewed the starting system geometry from the systems code. F. Najmabadi reaffirmed that the system code geometry output should only reflect the physics and engineering inputs and constraints. The project should question the validity of all the code inputs and not blindly accept the results. An example was the 20-30 cm thickness of the vacuum vessel proposed by Waganer, Lee, and Sviatoslavysky. (Action: Waganer to confirm VV thickness with due consideration for ports and support structures. EM loads on VV are assumed negligible.)
Leslie Bromberg showed a FEM analysis that allowed the TF magnet cap structure to be raised up vertically 1.0 meter from the previous position to allow for maintenance and plumbing. (D. Lee to revise his configuration to match.) With the reduction in cap coverage, the strains are increased, but the stresses are acceptable even with holes in the caps for vacuum ducts and structural supports. The method of coupling the TF coils to the cap structure is not explicitly modeled. To increase the (validity?) of the structural approach, it was recommended that some intercoil structure be added. Some questions were raised whether there is a location where the structure could be segmented. Leslie agreed to examine this.
Dennis Lee reviewed in detail the power core configuration that he evolved from the systems code data. Large horizontal ports through the vacuum vessel and cryostat enabled horizontal removal of almost all of the 1/16th outboard blanket and shield [A principal goal for maintenance is to provide straight radial extraction paths for all inboard and outboard modules.] The remaining thin wedge was determined by the location of the outboard TF coil legs and the thickness of the coil case, vacuum vessel, and clearance space. A clearance of 10 cm is recommended between TF coils and the VV for assembly and maintenance. C. Bathke will get information from Dennis to adjust the outer TF legs to enable removal of a complete blanket sector. Some uncertainty existed over the definition of the coils within the perimeter specified by the systems code. The discussion led to the resolution that the boundary of the coil includes the coil case, winding pack, and all internal structures. Dennis will remove any additional external structure representing a coil case. External support structures should be represented as separate parts
Considerable discussion time involved the location of the cryostat. The two options are (1) close- in and wrapped around the TF coils (ARIES concept), or (2) moved outside the vacuum vessel (ITER and TPX). The current design locates the cryostat outside the vacuum vessel, ala. approach (2). The rationale for this design choice should be documented, including the "pros and cons". Dennis showed one divertor vacuum duct on top, but another could be added on the bottom if required. Dai Kai will coordinate the vacuum system definition, including the external pump components, vacuum throughput analysis, and duct layout. Clement, Igor, and Dennis will provide support in defining the requirements (gas species and mass fluxes, configuration, and integration). D-K Sze will provide pipe sizes and entrance/exit locations for all blanket and shield modules to D. Lee for incorporation into the overall mechanical design. X. Wang will develop component blanket and shield design drawings.
Laila El-Guebaly discussed her analyses on damage and activation of the stabilizing coils. She also recommended replacing the U.S. ENDF/B-V data library with the new international FENDL library. [After the meeting, she conferred with Bathke, Tillack, and Sze and decided to use the FENDL library for any new analyses.]
D-K Sze mentioned that the results of the CaO coating test were negative. When placed in the ALEX loop, the effect from the CaO coating was absent. It was thought the temperature was not high enough to maintain the replenishment of the oxygen from the V-alloy to the CaO layer.
Fred Silady presented the design and performance data on the plate fin recuperators for use in the closed cycle power conversion systems of the GT-MHTGR design. They are much smaller than ordinary heat exchangers and are 95% efficient.
The Physics meeting concentrated on completing the Interim Report writing and outlining the priority physics necessary to support the RS design activity.
Interim Report: I. Chapter 6 1. Kessel/Bathke to prepare Figure 12 and add associated text 2. Jardin to complete references 3. Mau/Bathke to update LAR design point to STREAC17 4. Mau/Bathke/Ehst/Kessel to update RS in Table I to A=4 with wall at b/a=1.25 Add footnote on SS saying that old bootstrap model was used. Say in text that for SS, beta drops precipitously as wall is moved out, while for RS it doesn't. 5. Jardin to proofread and modify for submission to Fusion Technology. II. Section 8.3 ... Reversed Shear Physics 1. Kessel and Ehst to finish and provide to Miller III.Section 9.3 ... Low A Physics 1. Jardin to add paragraph on possible advantages of increasing kappa to 2.8, and to finish references and add summary 2. Kessel to rotate table 3. Mau to rewrite CD section RS Design Physics: 1. Kessel to redo baseline with kappa(x-point) about 1.9 so that the vertical stabilizing wall is behind the reflector, and with the kink-wall at 1.1 (as per Laila) 2. Kessel/Ehst to clarify tradeoff between edge CD and beta for A=4 RS 3. Bathke to perform POPCON analysis 4. TK to determine CD requirements for startup 5. TK/Ehst to look at how much rotation will be obtained from the RF 6. B.J.Lee to compute power and other requirements for a fake rotating shell (following Fitzpatrick/Jensen paper) 7. Kessel to compute vertical stability power requirements and to estimate how much cold structure will take place 8. TK to design RF launchers 9. Everyone to participate in divertor conference call
Action Items:All All inputs to Bathke for next strawman 2/21 Bathke Complete strawman w/outer TF legs increased 2/26 Mau Develop a non-inductive startup March Bathke Partial power operation POPCON analysis March Kessel/Ehst Tradeoff between edge CD and beta March Mau Design RF Launcher March Physics Group Complete Chapter 6 of interim report ?? Najmabadi Make presentation to DOE 2-3 wks Wong, El-G, Sze Summary of system choices 2/16 Steiner Provide preliminary safety analyses Next mtg Sze Determine b/a for 2nd V layer ASAP Kessel Analyze vertical stability of 3 cm W @650°, b/a=1.5 ?? Kessel Determine plasma elongation (k=1.9?) ?? Kessel Analyze kink stability of 2 cm V ?? Kessel Determine power/current for vert stab coil, Cu w/SS ?? D. Lee Revise configuration of power core, remove TF external cases March Waganer Analyze VV thickness 2/21 Wong Use ITER diffusivity, give Bathke impurity fraction in core, Bathke confirm magnitude of SOL increase, estimate heat flux including impurity’s radiation effect, TH & Stresses, vacuum pumping space 4 weeks or March Bromberg Incorporate intercoil structure, examine structure segmentation March Sze Provide piping data to El-Guebaly,D. Lee,X. Wang 2 weeks Sze, Sviat, Wong, Lee Dai Kai will coordinate the vacuum system definition, including the external pump components, vacuum throughput analysis, and duct layout. Clement, Igor and Dennis will provide support in defining the requirements (gas species and mass fluxes, configuration, and integration). March