ARIES Project Meeting Minutes
18 September 2013
Hampton Inn, Germantown, MD
Documented by M. Tillack
Ref: Agenda and Presentation Links: Meeting Agenda
Administrative and General Information
Welcome/Agenda - Mark Tillack called the meeting to order, explained logistics and summarized the agenda.
Presentation - On behalf of the Office of Science (signed by Pat Dehmer), Jim Van Dam presented a plaque to Prof. Najmabadi in appreciation of his many years of leadership in the ARIES program.
Opening remarks from OFES - Jim Van Dam discussed recent heightened interest in the cost of fusion power, which was triggered in part by an editorial in the New York Times by a Nobel Prize winner (Burton Richter). He expressed the sentiment that the ARIES Team is best equipped to respond to calls from the community to compare ARIES power plants with ITER, and requested that a small group within the Team provide a few pages for general consumption explaining at an overview level why we believe a future power plant would be cheaper than ITER.
He also felt that a separate effort may be warranted to scope out a possible technical activity to examine costs in more detail, for example for magnet systems. This prospective activity might explain individual cost elements, such as material unit costs, tooling, R&D, etc. in more detail. The question at hand in the near term is what it would take to perform such a study and what might be the usefulness and outcome (and NOT to perform such a study right now).
He requested a response to the first task within about a week. Les Waganer is taking the lead on preparing such a document, which will be reviewed by several Team members.
Review of ARIES ACT Project Status and Objectives - Farrokh Najmabadi noted that ACT-1 technical papers are mostly on track. After all papers are received and approved, Farrokh will forward them to Nermin Uckan for publishing in a special issue of Fusion Science and Technology Journal.
Our intent is to wrap up ACT2 at this meeting, with an intent to identify and prioritize remaining issues. We are anticipating some sort of closeout presentation(s) at DOE HQ in December, or possibly early 2014.
Al Opdenaker commented on the results of the peer review (see presentations), which were generally positive but criticized us on our outreach to the plasma physics community. Farrokh noted that we regularly try to obtain invited slots at APS and IAEA annual meetings, but lately the organizers have chosen not to include us. Jim Van Dam suggested that we consider a tutorial session on power plants at the next APS meeting.
ARIES ACT Task Results
Physics Analysis - Chuck Kessel reviewed the ACT2 design point and how we got there using systems analysis. Les Waganer asked whether ACT2 is a better basis for comparing to ITER. Chuck replied, "probably yes". But if we wanted to get a one to one comparison of ITER and power plant costs component-by-component, then we would have to make them identical in geometry, adopt similar power core technologies, etc.
Various ACT2 physics topics were covered, as shown in the presentation, on stability, current drive, TSC simulations, etc. The reference case for stability analysis uses the "no wall limit", but alternatives offering higher beta were also considered. We could show how some amount of wall stabilization helps by using the system code.
Alan Turnbull mentioned that we need to include stability coils and perhaps rotation. Chuck answered that we always have the RWM window coils in the design when we use wall stabilization (like ACT1). In ACT2 we do not have wall stabilization and would not need RWM coils, unless we assume some faraway wall is helping us. In that case we must put those coils back in again. The RWM coils are located behind the Ring Structure like the vertical position coils. In any case, we did not get analysis of the current and voltage requirements for the RWM coils, which would have to come from GA.
The ACT2 pedestal height is significantly higher than ACT1. Farrokh mentioned that pedestal height and beta-N have a large impact on the design point. Chuck indicated that he prefers to use EPED for extrapolation rather than experimental data. Farrokh asked why we have more confidence in EPED than experiments for ACT2. Alan Turnbull defended this approach, explaining that EPED has been shown to have very good predictive power over a range of machines. Chuck added that the problem with the semi-empirical formula is that it's a fit, with its own limitations with respect to extrapolation. Theory should have better predictive power.
In discussing the results, Chuck observed that the only way to lower the pedestal temp is to increase the density. Curt Bolton suggested we consider pellet injection in the edge for this purpose.
Chuck next reviewed ELM parameters. The temperature rise in the divertor is very high for large ELM's, making mitigation mandatory.
Ideas were discussed on how to use neutral beams. 1 MeV beams look possible for ACT2 (same energy as ITER), which makes them credible as an option (albeit an ugly option). Laila El-Guebaly suggested we will we need to design a beam dump. Chuck replied that hopefully by the time we have a power plant we would have adequate interlocks to prevent firing the beams without a plasma. Laila added that we need a lot of shielding around the beam lines. Chuck concluded by stating "at least we want to scope it out as an option".
Remaining tasks to be done include:
Farrokh commented that we ended up with a large machine because of the large number of constraints. It would be good to examine what happens if we relax one (e.g. Fundamenski). Chuck noted that the newest Fundamenski relation (about to be published, with Goldston as a co-author) is much worse than our assumptions. We chose the ITER formula because it's challenging, but not obscenely difficult.
Farrokh would like to see a table with parameter variations including higher beta-N, qdivob, n/nGr = 1.5, etc. We want to make the point that ACT2 has backed us into a corner, and changing any parameter tends to have a larger impact as compared with ACT1. Chuck thought this could be done using existing systems output.
Mark inquired how we will disseminate the results of ACT2. Farrokh requested that we collect whatever articles we can muster and then send them to FST again. If they amount to a full issue, then fine. If not, also fine. Mark recommended that someone write a separate article on systems analysis, since we didn't have one in the ACT1 special issue and there should be a lot of results when we consider all four ACT designs.
Engineering Analysis - Mark Tillack described the work performed over the past 4 months on ACT2 power core CAD, design and analysis. The analysis includes thermal, primary and secondary stress for a reference concept, and pressure stress analysis for an alternative concept.
Laila asked if it is possible in the reference design to locate the He manifolds behind the breeder rather than in the corners of the blanket as currently shown. Mark agreed to bring this suggestion back to Xueren and look for a solution that avoids nuclear hot spots.
Jake Blanchard felt that we might look at other boundary conditions in the 3D analysis, such as fixing rotations or forcing the top plane to remain planar. Mark noted that we are almost out of money and manpower will become difficult, so this may be too much to ask in the present study. (Later it was determined that Xueren had already examined a case with the top surface constrained to be planar, and saw little difference with the base case.)
Farrokh suggested that we could in principle omit the details of the cooling channels and perform 3D analysis of the full blanket, as was done for the vacuum vessel. Again, Mark pointed out that we do not have the resources to do these additional analyses.
Questions arose regarding support of the modules for the alternative small-module design. Mark pointed out that all of the small modules are combined into a single replacement sector, but they are mechanically independent. There are no sector side walls. Each module must contain the full pressure of the coolant independently.
After results of the 3 first wall design options were presented for the small-module design, it was clear that: (a) all of the options have too much steel, (b) option B is the best, and (c) there is still room for improvement by reducing the back wall thickness of option B. Mark agreed to provide Laila the optimized 2nd wall thickness that meets the stress requirement for 1.6 MPa coolant pressure, and Laila then agreed to look at the TBR prospects for this design using 1D or 2D analysis.
Disruption Analysis - Jake explained how they had added more details on the structures for the latest update of EM analysis. But the biggest impact on the results came from better characterization of the plasma. The inclusion of additional structures tended to reduce the currents and forces in the vacuum vessel.
Curt Bolton asked whether poloidal plasma currents were included in the analysis, as these could produce large effects. Jake explained that poloidal currents can flow in the structures, but the plasma was modeled with toroidal currents only. Chuck added that the disruption problem is complicated and we need to have an understanding of how faithfully we are modeling the problem. Jake's methodical progress is the right way to go, but we need to explain what is missing and what is future work. A dominant factor is the structure geometry: what is conducting and what is insulating, and connections of components to each other. In addition, halo currents (which we are not modeled here) can be very damaging.
The bottom line is that the loads are now low, and the location of the highest forces are inboard where the structure can handle it easily. We still need to look at force distributions on all components when the segmented blanket is included.
Farrokh mentioned that he wants a separate paper on the new vacuum vessel design and analysis, including disruptions.
Neutronics - Laila explained the various changes in the DCLL blanket analysis since 2011 that led to improvements in breeding. The TBR now varies strongly with enrichment. The current design provides a calculated TBR=1.05 with 40% Li-6 enrichment. Breeding in the top blanket does not contribute much to the TBR.
The rewelding limit was removed as a requirement on the structural ring. Without hot spots, everything outside of the blanket can be life of plant.
UW is planning to finish their ACT1 paper in a week or two, and then work on a paper for ACT2.
He cooling experiments - Minami Yoda presented a review and update on Georgia Tech experiments. Work is focused on the multiple jet "finger" design. The new helium loop is up and running with up to 10 MPa pressure and 295°C inlet temperature. Heating up to 2.7 MW/m2 is provided by an oxyacetylene torch. They are planning to loan an induction heater from INL to access up to 5 MW/m2.
This activity will be supported in the future by the US-Japan PHENIX collaboration. The technical thrust may change. For example, the Japanese are very interested in relaminarization (at low Re), which could lead to ineffective cooling.
Closing Remarks - Farrokh led a brief discussion of the status and plans. In 2 weeks he wants to finalize the schedule for our ACT1 special issue with FST.
Al Opdenaker asked if we could briefly summarize conclusions from the new studies (ACT1 and ACT2) as compared with the old (ARIES-AT is like ACT1, ARIES-1 is like ACT2). It would be useful to articulate what has changed in the past decade (or so) and what we have learned. Farrokh provided an impromptu summary.
We still need to look at ACT3 and ACT4 with the systems code.