Starlite Engineering Group Conference Call

13 Sept. 1995

The call-in phone number is 314-232-7776 and the time is 9:30 PDT (10:30 MDT, 11:30 CDT, 12:30 EDT).

C. Bathke, M. Billone, L. Bromberg, L. El-Guebaly, S. Herring, D. Lee, R. Miller, G. Sager, I. Sviatoslavsky, D.-K. Sze, M. Tillack, L. Waganer

note: action items are in bold-face.

The next call for the Engineering group will be Oct. 18th - same time and place. The Starlite full project conference call is scheduled for Sept. 27th, and the next project meeting will be Oct. 25-27.

Meetings and Reports.....

1. ANL design integration meeting

A. Summary of design integration activities (Waganer)

Minutes of the meeting were distributed electronically by D. K. Sze. These will be editted and distributed in final form. L. Waganer reviewed the main items discussed. One topic was the magnet design and its impact on maintenance. A TF crown structure is anticipated, making vertical maintenance unlikely. The strawman number of coils is 16, but this may change. The intercoil structure is being designed now.

V. D. Lee is working on the configuration and extraction scheme. Segmentation of the blanket and divertor was discussed. Ways to avoid full sector removal were examined.

A design which combines the vacuum vessel with the FW/B/S and divertor support structure was discussed. The cold shield would be attached to the vacuum vessel, rather than a separate "strongback".

Additional input is needed to define the divertor envelope. V. D. Lee needs further input from C. Wong to define the divertor slot depth, separatrix and X-point locations, etc. Additional input also is needed from L. Bromberg. These should be obtained prior to the IEEE meeting in order to make progress according to the Starlite schedule.

L. Bromberg reported that FEM work has been done to demonstrate static load support.

G. Sager pointed out that a memo was released on Aug. 29 by C. Wong with the initial divertor design philosophy. I. Sviatoslavsky and D. K. Sze noted that the design in the memo is simply a rescaled ITER design, and that the divertor configuration needs to be refined further in order to incorporate it into the Starlite design.

L. El-Guebaly reported some information from Physics group: the divertor slot will have ~50-cm poloidal width and 1-m depth. The depth is not measured from the X-point; it may start ~30 cm back.

V. D. Lee indicated that he needs hard coordinates. C. Bathke noted that a figure from the project meeting has the X-point indicated, such that we can guess the starting point of the divertor.

C. Bathke and L. El-Guebaly agreed that there is not enough room for shielding behind the divertor, especially on the inboard side. We need about a meter for the shield (80 cm is the minimum). We further assume there is no blanket at the back of the slot. It may be necessary to move the X-point and/or change the triangularity. Another solution may be to move plasma axis out. G. Sager said that the basis for the 1-m number is crude. We need 1-D calculations. (B. J. Lee has done only 0-D estimates) The slot depth could be made smaller.

C. Bathke will send e-mail to the Physics team asking for more info and informing them of the problem. He will try to get a reply soon.

It was suggested that V. D. Lee should work backwards from component sizes and calculate what is left for the divertor slot. We can give that number to the physics group to work on.

D. K. Sze suggested that we may need more coils to keep the TF ripple low if we pull in the coils. C. Bathke will examine variation of ripple with coil dimensions before the IEEE meeting.

R. Miller suggested we may want a number of coils compatible with 3 external loops. That means the number of coils would be a mutiple of 3. Going from 16->15 may not significantly degrade the ripple, but going from 16->18 may eliminate too much space for maintenance.

The idea of moving cold TF coils was discussed briefly. We should avoid any maintenance scheme which requires TF coild movement due to alignment problems, the large mass, and cryogenics complications.

B. Summary of meeting results and action items (Sze)

One of the topics not already discussed above is disruption effects. In ITER, Disruption effects are very large. We want to have an in-house capability to analyzse disruption forces. J. Blanchard (with help from Sze, Tillack and Hua) is assessing different codes, including CARIDDI and EDDYNET. EDDYCUFF is used by ITER and may be another possibility. J. Blanchard should contact Kessel to determine if the SPARK code (used by Bialek) could be used. It is coupled to the TSC code.

2. IEEE symposium

A. Status of symposium contributions (Tillack)

September 15th is the deadline for papers to be completed and distributed for project internal review.

B. Engineering group meeting agenda and arrangements (Sze)

We have a room Oct. 1 from 1-7 pm. The IEEE reception starts at 5 pm. Speakers should let D. K. Sze know about conflicts.

3. Status of FS and V reports (Tillack)

The FS report has been restructured. The two primary pieces missing from the report now are the material properties appendix and some new results on the cost comparison.

The radiation damage lifetime limits were discussed in depth. M. Billone can not find any compelling data which would clearly distinguish the lifetime of FS vs. V-alloy, assuming both maintain operating conditions within the specified design limits. Previously, a limit up to 200 dpa has been used for vanadium, whereas limits as low as 100 dpa have been used for FS. The lifetime of SiC is impossible to predict given the absent database.

Recommended wall load and surface heat flux limits were discussed. For FS, a peak surface heat flux of 0.8 MW/m2, corresponding to a peak neutron wall loading of 4 MW/m2, was suggested. For vanadium alloys, the recommended peak surface heat flux is 1.2 MW/m2 corresponding to a peak neutron wall loading of 6 MW/m2. The surface heat fluxes are based on an assumed wall thickness of 5 mm; higher limits may be achieved with thinner walls.

The systems code has dpa limits and wall load limits for each material. We need to assemble and document the recommendations for Starlite. M. Billone will send out e-mail on dpa and power density limits for project review. Bathke/Miller will incorporate the agreed-upon values into the systems code. For now, the systems code should use 15 MW-yr/m2 for both the lifetime of both FS and V.

R. Miller noted that these limits will have several impacts on the optimum point - larger machines, larger aspect ratio, etc.

M. Tillack asked if the availability is appropriately modified in the systems code based on the component lifetime and repair time. The availability module currently is not turned on in the code (all options are assigned a 76% plant capacity factor now.) This option will be turned on later.

The possibility that the divertor limits power density more than the FW/B was discussed. We should develop some kind of "limit algorithm" in the system code. R. Miller claimed that we've already accommodated this. It is more complex than purely engineering power density limits, and depends on physics and other considerations.

We need a divertor model from the Physicists to give us the best estimate of the divertor heat fluxes.

TASKS (status reports from task area leaders).....

4. Design Integration (Waganer)

R. Miller recommended that natural circulation should be considered in maintenance and configuration studies. D. K. Sze said that natural circulation has more impact for low-conductivity, non-conducting coolants, and may not be important for us. We do need to consider a dump tank for the Li.

L. El-Guebaly noted that the VV configuration currently shows 10 cm of solid SS. L. Waganer indicated that a 2-wall vessel is being considered. The VV design is still under development.

5. Maintenance approach (Lee)

D. Lee is still developing the modularization of the tokamak. The change-out times for the divertor and blanket also need to be characterized.

L. El-Guebaly ad D. K. Sze pointed out that the divertor has a lower neutron wall loading than the blanket, so in a low-erosion regime (which is probably mandatory), it could have a longer lifetime. One disruption per year has been assumed. Given these considerations, we feel we can make a good argument for the same maintenance schedule for both the divertor and blanket.

R. Miller reminded the group that the crown has to be removed to get the divertor out.

6. FW/B/S (Sze)

D. K. Sze has been holding discussions with individuals to initiate FW/B/S tasks. Completing writing assignments is also a high priority.

7. VV/Primary Structure (Sviatoslavsky)

The ITER design methodology has been examined. The current proposal for Starlite is to use the VV to support internal components, rather than to include a separate strongback. The design concept under review uses 3 modules per sector for the outboard blanket. Segmentation leads to less translation required for maintenance. Inboard there will be only one module, brought out radially with no translation. The divertor probably needs segmentation as well. The preferred approach is to attach the divertor to blanket somehow. The shield also must be removed between coils.

C. Bathke pointed out that in the past there was a cost penalty accepted for single-piece maintenance by pushing the coils out (to become larger). He asked if anyone has looked at the trade-off. I. Sviatoslavsky indicated he has not performed the trade-off, but suspects this would be too costly.

However, this was the strategy used in ARIES, and is the baseline for the current system code. The additional cost was estimated as 3-4 mill/kW-hr. The additional time needed for maintenance of segmented pieces has to be considered.

Another consequence of full-sector maintenance is the support structure. There is not much VV left if a single-unit sector has to be removed. I. Sviatoslavsky will look at this again.

M. Tillack asked if we can we quantify the difference in MTTR depending on segmentation. I. Sviatoslavsky answered that we don't really know the difference in MTTR. In ARIES there was a desire to avoid compound motions. R. Miller added that we can try to quantify these things in the availability subroutine.

L. Waganer added that plumbing difficulty should be a factor in the trade-off.

M. Tillack asked if the configuration is taking account of thermal stresses caused by a cold zone at the back of the blanket. Currently, the design calls for the VV and secondary shield to run at 300 C. The primary shield and FW run at 600 C. Between the two, we anticipate a flexible connection. Vertically, the blanket is fixed at the midplane, and a moveable joint exsists at the top and bottom.

D. Lee asked if a hot VV is a problem for coil temperatures. L. Bromberg answered that you need only a few cm more insulation.

R. Miller and I. Sviatoslavsky agreed that the real problem with a hot VV is startup.

8. PFC (Sager for Wong)

G. Sager noted that a memo was sent out 2 weeks ago on the divertor design.

The issue of plasma impurity requirements was raised. G. Sager said that there is not a significant cost penalty for plasma impurity. The costing depends more on current drive efficiency.

C. Bathke suggested that Zeff doesn't really affect the divertor heat loads (maybe only a 25% effect).

D. Sze raised the point that with a gaseous divertor (probably Ar with higher melting point) you need high-temperature regeneration in cryopumps. Scott Willms will help analyze this. The low temperature part of the cycle is defined by H and He. The energy needs for this cycle may be prohibitive.

9. Neutronics (El-Guebaly)

H. Khater is performing activation calculations for divertor cassettes. The useof 2-um to 2 mm of W coating is being assessed.

D. Sze summarized analysis of activation of Ar (which activates to K). H. Khater did some calculations and found a fairly small production rate.

10. Magnet Systems (Bromberg)

Al stabilizer in the magnet conductor was examined. It doesn't look too good.

New costing numbers were generated for the systems code. R. Miller/C. Bathke will incorporate this into code.

11. Safety (Herring)

The safety standard should be going out for broader public review (~300 people).

R. Miller asked if INEL is plugged into the hazard analysis work. S. Herring answered "In general, yes. But not in the last couple of weeks."

R. Miller added that they are analyzing ARIES designs. We need to know if they find any obvious flaws to guide the Starlite design process.

They are trying to put more realism into hazard analysis. L. El-Guebaly and R. Miller are working on waste streams and implications.

Final comments:

D. Sze asked if the cost of WC has been resolved. R. Miller is still working on it. It isn't settled yet.

D. Sze reminded the group about e-mail sent on tritium accountability.