SRF Cost Reduction for Project X Bob Kephart

SRF Challenges for Project X Adoption of a 3 GeV CW linac followed by a 3-8 GeV pulsed linac for Project X results in a very powerful intensity frontier accelerator complex… but presents new challenges  Needs six different cavities optimized for changing velocity (  ) of Protons  Four different frequencies (162.5, 325, 650, 1300 MHz)  Five of these cavities are completely new for Project X (vs 2 for SNS, 1 for CBEAF)  Requires development of seven different styles of cryomodules “flawless execution” of a ~ $ 1 B class DOE project based on these cavities and CM requires development of representative prototypes such that performance and costs are well understood Requires a major R&D effort just to achieve the desired machine performance, reliability, etc However, COST may drive if/when PX will be constructed RD Kephart FNAL PX cost reduction Sep 132

Project X Accelerator Costs Estimate – Full Scope thru 8 GeV Comparison only! 2010 cost estimate, direct $ (no overheads), no contingency, etc. Cavities and CM represent 31% of the Project X Costs ~ $500 M with OH & contingency Cryogenics = 14% of Project Cost ~ $ 200 M with OH & contingency Cryogenics ~ 80% is CW linac RD Kephart FNAL PX cost reduction Sep % 10% 11% 3%

MEBT SRF Map for Project X HWR SSR1SSR2  =0.6  = MeV 1.3GHz ILC 3-8 GeV GeV RFQ H - LEBT RT (~15m) Pulsed CW RD Kephart FNAL PX cost reduction Sep 13 * 5 warm and 5 SC doublets. ** All doublets and correctors are warm MeV SRF part of PXIE Project Cost Risk Non-PXIE = 451 cavities/ 62 cryomodules 4 4

Project X Cryomodule Cost Estimate – Full Scope thru 8 GeV While important technically, cavities and CM being developed for PXIE represent only 7% of ~ $500 M Project X Cavity and CM Costs Comparison only! 2010 cost estimate, direct $ (no overheads), no contingency, etc. PXIE 650 MHz CM represent more than half the costs and dominate the Project X cryogenic heat load 1300 MHz CM are also a big piece of the costs RD Kephart FNAL PX cost reduction Sep 135 3% 4%

Driving the costs down: Current R&D activities  Project X: Cavity and Cryomodule Development (1 st articles)  HWR (ANL)  325 MHz SSR  650 MHz Elliptical (LE and HE)  1300 MHz: Elliptical, beta = 1, pulsed  Development of U.S. cavity and CM parts industrial base  Infrastructure and performance testing of 1 st articles  The current PX R&D program is largely pointed at design validation and technical improvements vs cost reduction  Why? Funding has been limited  focused our R&D efforts on tasks which lower technical risk and validate costs… ie the absolute minimum tasks to launch a project Exception  SRF materials program (Cavity Qo, cavity processing & yields)  Low hanging fruit (factor of two in Q0 saves 10’s of $ M in cryo RD Kephart FNAL PX cost reduction Sep 136

Project X Cavity and CM development Lots of cavity and cryomodule development remains to lower Project technical risks and validate costs (even if those costs are higher than achievable) RD Kephart FNAL PX cost reduction Sep 137 PXIE

8 New surface treatments developed at FNAL for higher Q (Slides from Anna Grasselino) 16 2K, 1.3 GHz Demonstrated on several cavities Improvement over baseline, component improved Maximum Q achievable with ideal cooldown (16MV/m, 2K, 1.3GHz) EP+120C (standard ILC) e10--~2.2e10 EP + 120C + HF rinse~2.3e10~30-40%, residual3e10 EP + Annealing + no chemistry ~2.7e10~50-60%, residual3e10 EP + Gas (nitrogen or argon) bake + EP ~4e10~ %, BCS and residual 6.75e10 Golden opportunities!

9 Simple higher Q 0 recipe: 120C bake +1 HF rinse Single HF rinse (5 min) followed by water rinse is beneficial for the medium field Q value – gains of up to 35% measured at 70 mT A. Romanenko et al, Phys.Rev.ST Accel.Beams 16 (2013) T=2K 30% 35% 1.3 GHz, 2K

10 Annealing as a last processing step A.Grassellino et al, GHz, 2K 50-60% EP + 800C 2 hrs micron EP + 120C Systematically low R 0 ~ 1nΩ Extra cost savings from skipping the post furnace chemical processing higher Q

11 Nitrogen or argon baking: world record Q values A.Grassellino et al, 2013 Supercond. Sci. Technol % Gas bake plus EP: state of the art treatment for highest mid field, reverses the mid field Q-slope! Technology ready at FNAL: success rate 100%, realized on ~ ten cavities, also on 9 cell cavity. Cornell working on reproducing the recipe Quench systematically 20.3 MV/m (86 mT) – large enough margin for 16MV/m 2K (8 to 1.8K) ~7e10! 1.3 GHz, 2K

12 Q preservation in a cryomodule: cooldown requirements Q is determined by the surface processing, in a cryomodule at best it will be maintained – Fast cooldown through hydrides formation region – found to be 175K-90K – Slow and T-homogenous cooldown through Tc to minimize trapped magnetic flux due to thermocurrents – Some modifications to cryomodule and vessel might be needed for high Q

13 Strategy for Px cost reduction via Q 0 Fermilab has developed world leading expertise in surface processing for high Q 0 Techniques well grounded in SRF surface science Nitrogen or argon bake shows great promise for producing Q 0 above 3e10 at 16 MV/m However, R&D is required to build statistics and move this R&D from single cell tests to multi-cell mass production Also need to study preservation of Q 0 in a cryomodule

Apply Basic Rules of Project Cost Reduction Focus on the Cost Drivers for SRF this is Cavities, Cryomodules, and Cryogenic systems Cost Reduction by Design 1) Reduce or eliminate entire systems - o Higher Q 0  smaller cryogenic systems o Active microphonic control vs RF power and bandwidth? ….higher Qext simplifies microphonics but costs in RF power o Cavity shapes that reduce FE and multi-pactor losses o Choice of operating gradient (# CM’s vs cryo plant and civil) 2) Insure that spec’s do not exceed requirements o Understand cost of alignment tolerance specified o Pressure ratings: Do our standards make sense? o Cost of thermal “safety factors” count it only once! o Instrumentation; is it really required, or just nice? RD Kephart FNAL PX cost reduction Sep 1314

Apply Basic Rules of Project Cost Reduction Cost Reduction by Design 3) Reduce the parts count ! o Cryomodules are expensive with lots of part o Many assembly steps, lots of labor o Are all these parts and steps required ? 4) Eliminate expensive materials or process steps o Reduce amount of Nb (e.g. Nb films on Cu cavities o Eliminate Titanium parts, vacuum feed thrus, etc. o Eliminate machining & EB welding of cavities via...hydro-forming, and perhaps laser welding? o Optimize time and labor for CBP and move bulk …Niobium removal step to industry ? o Eliminate test of dressed cavities (reliable assembly) RD Kephart FNAL PX cost reduction Sep 1315

Apply Basic Rules of Project Cost Reduction Cost Reduction by Design 5) Design for manufacturability o Labor is expensive! o Parts that are easier to make o Parts that snap or lock together vs require alignment 6) “Total Cost” analysis o Reduce the “Total Cost” of parts and assembly labor o Include yield and rework in cost analysis o Design for increased labor productivity and lower waste o Include the type of labor required in the analysis. A more expensive part snapped into place by less skilled worker may cost less than a cheaper part that requires a skilled technician to install RD Kephart FNAL PX cost reduction Sep 1316

Apply Basic Rules of Project Cost Reduction Cost Reduction by Standardization 1) In general, Standard parts are cheaper than custom parts 2) Sharing parts with other projects can significantly lower costs for both Projects (or especially for the 2 nd project) o Lowers design cost o Lowers learning curve and avoids costly mistakes o Established industrial component sources o e.g. leverage overlaps with ILC R&D, XFEL, LCLS-II 3) Usually we design for the “Performance and Upgrade” optimum… vs standard parts or the “cost optimum” o This approach is deeply ingrained in our culture o However if our Projects are NOT approved because cost are too high, we may wish to examine this approach RD Kephart FNAL PX cost reduction Sep 1317

Specific Recommendations Near-term R&D 1) High Q 0 R&D with the goal of Q 0 = 3-4 x in 650 MHz multi-cell cavities in a cryomodule 2) Review specifications of all PX SRF components to insure they are not over specified 3) Exploit possible synergies with LCLS-II and other projects Mid-term R&D 1) Work with industry to design cavities and CM for manufacturability and reduced “total costs” Long-term R&D Goal: 1) Spun or hydro-formed 650 Mhz copper cavities with mirror smooth CBP surfaces, coated with Nb3Sn films achieving high Q 0, while operated at 4K in Stainless He vessels in a cryomodule designed for low total cost (low parts count and simplified assembly) RD Kephart FNAL PX cost reduction Sep 1318

Conclusions The components that make up the SRF linac in Project X are currently complicated, expensive, and drive the overall costs PX R&D is focused on 1 st article production, technical demonstrations, and establishing a cost basis for a Project Standard Cost Reduction methods could lower costs but will take time and money to pursue… industrial partners could help RD Kephart FNAL PX cost reduction Sep 1319