ATW accelerator technology in US roadmap

Progress in Nuclear Energy, Vol. 38, No. 1-2, pp. 25-64, 2001 © 2001 Published by Elsevier Science Ltd. All rights reserved

Pergamon

Printed in Great Britain

0149-1970/01/$ - see front matter

www.elsevier.com/locate/pnucene PII: S0149-1970(00)00095-0

ATW Accelerator Technology in US Roadmap G. Lawrence, K.C. Chan, S. Schriber, T. Wangler (Los Alamos National Laboratory), M. Allen (Stanford Linear Accelerator Center), W. Funk (Westinghouse Savannah River Company), T. Meyer (Advanced Energy Systems, Inc.), R. Noble (Fermi National Accelerator Laboratory), K. Shepard (Argonne National Laboratory), D. Shipler (Pacific Northwest National Laboratory), P. Wanderer (Brookhaven National Laboratory), and T. Ward (Department of Energy).

1.0 Introduction and Background Development of the 1999 ATW (Accelerator Transmutation of Waste) Roadmap was initially driven by a specific reference ATW deployment scenario that could handle the projected 87,000 tonnes of spent fuel produced by existing US commercial power reactors during their service life. This reference deployment scenario permitted development of total life-cycle costs and an assessment of the ultimate scope of a national nuclear waste transmutation program. Within the Roadmap effort, an Accelerator Technical Working Group (ATWG) was charged with identifying the accelerator technical and performance issues that must be resolved before deploying such an ATW system in the United States (using the reference system concept as the framework), and with formulating an appropriate R&D plan. The efforts of the ATWG were integrated with those of two other TWGs (covering Target/Blanket technology and Separations technology) by a Systems Integration Working Group. A Steering Committee provided general guidance. The specific ATW deployment scenario is one in which eight large transmutation stations would be built in the US, each containing eight subcritical fission assemblies driven in groups of four by two high-power proton linear accelerators. Construction and operation of full-scale ATW plants would follow a prototyping program in which an ATW demonstration facility (DEMO) would be built and operated at steadily increasing transmuter power levels. As the Roadmap evolved, it became clear that a preferable approach to development of ATW technology would be science-driven rather than deployment driven, and that the focus should be on an initial 5-year R&D program. This program would address the main ATW technical issues, derive system requirements, conduct trade studies to compare options, make initial technology selections, and develop a preconceptual design for an ATW demonstration facility, and by extension for the ATW plants. The 1999 ATW Roadmap Report [1] transmitted to Congress makes that recommendation. The accelerator technology R&D roadmap [2] developed by the ATWG is described in this paper within the above evolutionary context. The program, while influenced by the specific reference ATW Roadmap deployment scenario, would broadly address the key accelerator technical issues, begin development of needed accelerator technology, carry out trade studies to compare accelerator architecture and subsystem options, and produce pre-conceptual designs for both an ATW plant accelerator and the accelerator in an ATW 25

26

G. L a w r e n c e et al.

demonstration facility. Estimated costs and schedule of such a program are provided in the Appendix to the Roadmap Report to Congress.

1.1 Reference ATW Plant Accelerator

The reference ATW plant concept that formed the framework for the US ATW Roadmap process is shown in Fig. 1.1 Two 45-mA, 1-GeV proton linear accelerators (linacs) would each provide beam to four subcritical fission assemblies, with the latter sized at 840 MW m. At an assumed thermal-electric efficiency of 0.38, these transmuters produce a total of 2555 MWe of electricity. About 380 MWe is used to operate the plant, and the balance of the electric power (about 2175 MWe) is exported to the grid. A linac is the baseline for the US Roadmap because of the high proton power needed to drive each system. For other ATW reference plant concepts in which the needed beam power is substantially lower, an advanced cyclotron might provide an alternative accelerator option. The beam energy is chosen at 1-GeV to obtain efficient conversion of beam power to spallation-neutron production, and the beam power is chosen to drive the transmuters at their minimum estimated neutron multiplication factor. Two stages of radiofrequency (RF) beam splitters subdivide the current from each accelerator equally into four separate CW beams that are delivered to the subcritical fission assemblies. The beam transport system would be configured so that each set of transmuters can be driven by either accelerator, and the transmuters are connected to steam generators in pairs. When one linac is out of operation, or transmuter refueling is taking place, all steam turbines would be turning and at least half of the electric power capability of the plant is being exported to the grid. This accelerator and beam-distribution architecture greatly increases the overall reliability of the plant as a whole, in terms of electric power generation. Other accelerator and beam transport geometries are possible and will be evaluated during the 5-year initial ATW program. Iniector I NC Lina¢

SC Linac

BeamChopper

1000MeV

RF Splitter

• High-gradient SC main linacs with injection from 10-MeV NC linacs. • Beam power is distributed equally to four production transmuters using RF splitters. • Transfer lines allow each linac to supply beam

22 5 mA

( Dipole Transmuters

\

TransferLines- -

840 MWth

/~ ~ D

to either set of transmuters.

' /~ O.>\ O

• Beam to individual transmuters can be inhibited by RF choppers in linac front ends, • Generators connected to transmuters in pairs to assure continuity of electric power export. 45 mA

Fig. 1.1 Reference APT plant concept, with two 45-MW linacs and eight transmuters.

ATW accelerator technology in US roadmap

27

The reference accelerator design is derived from the high-power CW proton linac [3-5] proposed for the Accelerator Production of Tritium (APT) project, depicted in Fig. 1.2 below, but there are differences because of increased emphasis in the ATW application on beam continuity (reliability), power efficiency, and the need for delivering beams to multiple targets. The APT linac, which has an output energy of 1030 MeV and a CW beam current of 100 MeV, is a hybrid normal-conducting (NC) - superconducting (NC) machine. The low-energy section to 211 MeV is made up of water-cooled copper accelerating structures. A proton injector is followed by a 6.7-MeV 350-MHz RFQ (radiofrequency quadrupole), which feeds a 97-MeV 700-MHz CCDTL (coupled-cavity drift-tube linac). This section is completed by a 211-MeV CCL (coupled-cavity linac) also operating at 700 MHz. The high-energy portion of the linac is composed of 5-cell 700-MHz elliptical niobium super-conducting cavities, which are cooled by liquid helium at 2.15K and deployed in cryomodules containing two, three, or four units [6]. Two cavity design beta values (0.64 and 0.82) serve to efficiently span the energy range from 211 MeV to 1030 MeV. NC Linac 350 MHz l l 75 keV 6.7 MeV

700 MHz i ~ ~~ .....

l 97 MeV

SC Linac 700 MHz i~"~ I .... 4"3"5"5Mvtrn 211 MeV 749 m

l!:~4~-~i~'ii~ ~ ' ~ ] 471 MeV

i 1O0 mA

s'25 Mv/m 1030 MeV m-

Fig. 1.2. Baseline design for APT linac

A possible architecture for an ATW reference plant linac is shown below in Fig. 1.3, on a similar length scale to that in Fig 1.2. As in the APT linac design, it consists of a lowenergy NC linac injecting into a high-energy SC linac. However, this ATW strawman linac design calls for SC cavities with accelerating gradients about a factor of two higher than in the APT linac in the high-energy region of the accelerator, and for making the transition from NC to SC accelerating structures at a much lower energy than in the APT linac. Thus almost the entire ATW linac would be made up of superconducting RF cavities, and the total linac length is reduced by more than a factor of two in comparison with the APT machine. The high-beta section, above 78 MeV, would employ elliptical multicell cavities similar to those in the APT superconducting linac. The low-beta section, from 11 MeV to 78 MeV, would employ spoke-loaded 1/2-wave resonators, similar to those proposed for the Rare Isotope Accelerator [7] by Argonne National Laboratory. This brings the advantages of larger beam aperture, higher gradients, and potentially higher reliability to this part of the accelerator. In principle, it is possible that this kind of SC cavity technology could be employed immediately following the RFQ.

G. L a w r e n c e et al.

28

NC Linac

SC Linac

350 MHz q

700 MHz

= 45 mA

,E l ....... i nlector ......... 11

33

K 78

4

t

I

211MeV 390MeV

I IO00MeV

296 m

Figure 1.3. Reference accelerator concept for the ATW plant.

1.2 ATW Accelerator R&D Roadmap During the first five years of a science-based ATW development program, accelerator work would be focused on: • Addressing the beam continuity issue through 1) failure analysis of components and subsystems, 2) a program to develop and test high-reliability equipment and fast beam recovery schemes, 3) accelerator architecture trade studies. • Developing and testing key components that could have a high impact on the plant (and demonstrator accelerator) designs and that address performance requirements and design selections, such as high-gradient SC accelerating cavities, and RF beam splitters. • Putting together a preconceptual design for an ATW demonstration facility accelerator. This would be accomplished during the last two years of the initial R&D phase Early in the 5-year program, trade studies would address the overall ATW plant architecture, especially the issue of whether multiple transmuters should be driven by a single high-power accelerator or each transmuter should be supplied by its own (lower power) accelerator. Factors such as capital and operating costs will be balanced against reliability, availability, and operability. The outcome of these studies may, of course, significantly impact the direction of the accelerator R&D program. The manner in which the 5-year R&D program would fit into the Roadmap-based plan for deployment and testing of an ATW demonstration facility is summarized in Fig. 1.4. below. Following completion of a preconceptual design, and assuming Congressional approval to proceed to the demonstration stage, a conceptual design would be developed, along with a cost estimate and construction schedule. The standard project phases would then proceed, including preliminary and final design, construction of the demo accelerator in its initial power configuration, followed by beam commissioning and reliability studies. There would then be a several-year operating period during which transmuters of increasing capacity would be built and tested. After this time, the accelerator would be upgraded to full beam power by adding RF systems. This full power system would be operated for several years to demonstrate integrated ATW technologies under prototypical plant conditions, prior to making a decision to begin deploying full-scale waste-transmutation plants.

ATW accelerator technology in US roadmap O0 01 02 03 04 05 06 07 08 09 10 l"f

29

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

Fig. 1.4. Implementation schedule for linac in ATW demonstration facility.

1.2.1 Beam Reliability The overall availability of the present generation of large accelerators that are used to serve nuclear physics, high-energy-physics, or materials-research programs is very high, generally exceeding 85% of the scheduled running time; the next generation of accelerators (SNS, APT) is being designed with even greater attention to availability. However, for ATW systems, in which the accelerator is driving a subcritical assembly and electricity is being exported to the power grid, there are issues that go beyond the requirement of simply a high availability. For the ATW application, continuity of the beam is very important, with the spectrum of beam trips or interrupts being a matter of significant concern. Repeated power transients in the transmuters (caused by brief beam interrupts) could cause damage to fuel and structural components, impacting useful lifetimes of the subcritical assemblies. Longer beam interrupts could produce step changes in exported electricity that would make the power produced by ATW systems less valued than that from more reliable sources. The critical determinants are the duration and frequency of beam interrupts, and the thermal response times of the driven assemblies and power-generating systems. Beam trips shorter than the thermal response times of the transmuter (100 ms - a few sec) would cause relatively low mechanical stresses, since the induced temperature swings are modest; thus even a relatively large number of such short beam trips would be tolerable. Longer interrupts would be more serious in terms of cyclic stress damage. The response time of the electric-power generating system is longer, ranging from tens of seconds to several minutes, so it can accommodate moderate-length beam trips. However, even a relatively small number of longer outages per year would negatively affect the price for which ATW-generated electric power could be sold. The need for very high beam reliability in ATW systems is a new and clearly demanding performance requirement for high-power accelerators; how far we need to go is not obvious and will require considerable study. One standard of comparison is that of

30

G. L a w r e n c e et al.

modern commercial power reactors, which have only a few unplanned power trips per year. Modern research accelerators, in contrast, typically have a large number of shortduration beam trips (of order one per hour, or about 104 per year). In general, this behavior causes no special difficulty for users, since the normal production criterion is the total integrated current on target per unit time. Design priority has been given to protecting the accelerator and its subsystems from damage, while achieving a high average production availability. Beam continuity, i.e. the absence of beam interrupts, has heretofore been of much lower importance as a design criterion. At first glance, one might think that accelerator designers are faced with a challenge on the order of a factor of 1000. However, this is not the case, since more than 90% of beam trips are in the very short (< 1 second) category, and thus have little or no impact. The remaining 5-10% of beam trips (greater than a few seconds) are the ones that must be dealt with through equipment reliability improvement, and in the accelerator architecture and design.

1.2.1.1 Reliability Analysis and Testing The beam reliability issue would be addressed both through linac architecture trade studies and analysis, and a multi-faceted R&D program involving the development of prototype high-reliability equipment. These programs would be aimed at achieving a frequency of beam interrupts orders of magnitude lower than that which exists in today' s generation of accelerators. The program would initially identify the causes of unreliability or failure in the equipment systems that affect beam acceleration and delivery to the transmuters. The critical equipment systems would be redesigned and rebuilt as high-reliability prototypes to eliminate causes of failure; these would then be tested intensively and extensively to verify failure-free performance. High-reliability linac design architectures would be developed, and assessed in terms of tolerance for individual equipment element failures. These architectures would contain improved equipment redundancy, capability of ultra-fast compensation for failed accelerating units, and greatly reduced rates of equipment failure from all sources. Reliability analysis and linac architecture studies would begin in the first year of the program and continue for up to three years. Equipment reliability testing begins in the second year of the program and would run to the end of the 5-year R&D period. A combination of existing test facilities and purpose-built reliability evaluation and demonstration test stands will be used to carry out the component and operating-systems reliability test program. This program would involve testing and improvement of existing systems, such as the LEDA front-end accelerator prototype [8] and the medium-beta cryomodule prototype being built for APT [9], as well as the building and testing of new components and subsystems designed specifically to operate with greatly reduced failure rates. The results of these testing programs would be used in design of the ATW demonstration accelerator. Reliability testing would continue throughout the operation of the demo accelerator, using this machine as an long-term test bed to evaluate final choices for an ATW plant accelerator design.

ATW accelerator technology in US roadmap

31

1.2.2 Development of Key Linac Components: Issues 1.2.2.1 High-Gradient Superconducting RF Cavities For the superconducting radiofrequency (SCRF) cavity R&D program, the most important technical issues involve demonstrating reliable and reproducible attainment of high accelerating gradients in the elliptical multicell cavities used for the high-beta linac, simultaneously with a high cavity-quality factor (Q). The former is needed to minimize the length and cost of the linac, and the latter is needed to minimize the heat load to the plant cryogenic system. Most of the accelerator would be constructed from niobium superconducting RF cavities operated at temperatures between 2.1 K and 4.2 K. This maximizes the efficiency by which RF power is converted to beam power, essentially eliminating RF losses in the cavity walls and reducing the number of RF power systems required to drive the beam. Because very high accelerating gradients can be achieved in SCRF cavities, the linac length is kept to a minimum. The combined effects of reduced length and lower RF power minimize accelerator capital costs. Eliminating RF wall losses minimizes the power that must be recycled to the plant to operate the accelerator, and thus reduces operating costs. The reference linac design in the ATW Roadmap calls for SCRF cavities with 10 MV/m gradients in the high-velocity region of the accelerator, which is within the current state of the art [9] but twice that used in the APT linac design [3]. New types of SC accelerating structures (spoke resonators) [7] are being developed for the low velocity region (13 < 0.5), which provides the potential in an ATW linac for extending the SC cavity technology down to much lower energy than in the APT linac. The additional gains in power efficiency and reduced length for this part of the linac (in comparison with NC accelerating structures) are significant. An additional improvement over the APT linac design is larger beam apertures that can be achieved, which would reduce beam losses and improve operability in the low-beta part section.

1.2.2.2 High-Efficiency R F Generators For development of advanced RF generators, the main challenge is to achieve highefficiency conversion of DC power to RF power in a simple, reliable, and low-cost device. This hardware development activity has the greatest promise, after high-gradient SC cavities, for lowering accelerator capital and operating costs. The reference design of the ATW linac RF power system is based on the use of high-power (MW-class) klystrons, as in the APT linac. However, inductive-output tubes (IOTs) have considerably higher efficiency and are more compact. The APT program has been developing a high-order-mode IOT 1-MW prototype at 700 MHz. This tube shows promise, but its design is fairly complex, and it may fall short in terms of reliability. Multi-beam klystrons are another option for the ATW RF generator.

32

G. L a w r e n c e et al.

1.2.2.3 R F Beam Splitters For development of RF beam splitters, performance issues involve the level of deflection field that can be achieved reliably in the SCRF cavities and whether the desired deflection angles can be achieved without beam loss. The first issue can be addressed by testing the splitter cavities with RF power alone; the second requires a beam test, which could be carried out either on LEDA or on the LANSCE linac, both at Los Alamos. The ATW plant reference concept calls for subdividing the continuous wave (CW) beam from each high-power linac into four equal-intensity CW beams, using RF beam splitters. These beam splitters would be constructed of sequences of SCRF cavities supplying deflection (transverse) fields instead of accelerating fields. Similar devices have been built and tested for particle separation in high-energy beams, and the ATW beam splitting requirements appear practical. The beam splitter is the major component that must be demonstrated to confirm the feasibility of producing multiple CW beams for driving several transmuters, using a single accelerator.

1.2.3 Pre-Conceptual Accelerator Design During the last two years of the 5-year R&D program, a pre-conceptual design would be developed for the accelerator and beam transport system for a proposed ATW demonstration facility. The beam power specifications and staging of this accelerator would result from trade studies and system analyses to be performed during the first two years of the program. The pre-conceptual design would define the main features of the demonstration accelerator performance and architecture, provide a basis for conceptual design, outline additional R&D programs needed to reduce risks in the design, and provide a preliminary cost estimate.

2.0 Accelerator Technology Options; Linacs and Cyclotrons The two accelerator types that are under consideration for high-power medium-energy applications worldwide are CW RF linacs and CW cyclotrons. In the US Roadmap, because of beam requirements in the 40 - 100 MW range for each plant, the emphasis has been on linacs. Cyclotrons could play a role if the beam requirements for each ATW system were substantially lower, in the 10 - 20 MW class. There are presently no CW high-power proton linacs in operation. However, mature and well-reviewed designs for high-power linacs have been developed during the past few years at several laboratories, like the one for APT [3], and show that there are no fundamental obstacles to reaching 100 MW or more of beam power with this kind of accelerator. That conclusion is supported by extrapolation from the highest-power existing pulsed linac (the 800-MeV, 17-mA peak-current, 6%-duty LANSCE linac at Los

33

ATW accelerator technology in US roadmap

Alamos). In general, the CW proton linac seems well-matched to the power and energy requirements of the plant outlined in the ATW Roadmap (1 GeV energy and 45 mA beam current). Depending on cost, it may be practical to build the ATW linacs at half this power level (22.5 mA beam), serving only two transmuters with each accelerator. This would significantly improve the reliability and flexibility of the ATW plant. The highest power proton cyclotron is the 590-MeV separated-sector two-stage unit at the Paul Scherrer Institute (PSI) near Zurich, which presently operates at 1.7 mA CW beam current; there are plans to upgrade it to 2.5 mA. Preconceptual designs of advanced versions of this kind of machine have been put forward for up to 10 mA at 1 GeV, notably by the PSI group [ 10,11 ]. These designs are generally based on extrapolations from the PSI system design, and involve a similar cascade of two cyclotrons, injected by a low-energy accelerator. The PSI concept for a 1-GeV final-stage ring cyclotron is shown in Fig.2.2. race

Key Parameters Output Energy Injection Energy Output Current Max. B Field Cyclotron Freqency Accel.. Frequency Peak Voltage Extr. Turn Separation No. of Turns

1 GeV 120 MeV 10 mA 2.1 T 7.36 MHz 44.17 MHz 1 MV 9 mm 140

Magn~est

"~/~

Sector

Ac~elvetriat:ng" ~ . ) < ~

Cavities

"x~/"

~/

Beam from 120-MeV Cyclotron

Fig.2.2. I-GeV 10-mA ring cyclotron proposed by PSI for ATW service. This machine is injected by a 120-MeV separated-sector cyclotron.

The CW cyclotron appears to be a contender in the beam power range up to 10 MW, although there may be significant technical challenges in reaching that level. Going beyond that in a single cyclotron system seems problematical. A recent workshop investigating the prospects for designing and building high power cyclotrons concluded that there are serious beam-dynamics constraints that would prevent going to significantly higher currents and energies [12]. As the beam current increases above a few milliamperes, space charge forces begin to overcome the relatively weak focusing forces available in this kind of accelerator. Also, as the proton energy and current increase, orbits at the cyclotron periphery become too tightly packed to permit low-loss extraction.

34

G. L a w r e n c e et al.

Capital-and-operating-cost advantages presently appear to favor an ATW plant architecture that has one or two large high-power accelerators providing beam to multiple transmuters, as opposed to one in which each transmuter is driven by its own lowerpower accelerator. This leads to a high-power linac being the chosen baseline accelerator driver for the reference ATW Roadmap concept. However, if other factors, such as the possible requirement for intensity control of the beam currents driving individual transmuters, become more important, and/or if costs per unit of beam power can be brought down for lower-power accelerators, then this picture could change. An objective trade study and cost analysis is planned for early in the 5-year ATW R&D program that will compare cyclotrons and linacs over the relevant ATW beam-power range, in terms of cost, reliability, and other performance factors.

3.0 Accelerator Requirements for the DEMO Facility and the ATW Plant To fit in with the deployment plan for an ATW demonstration facility (DEMO), as described in the Appendix of the ATW Roadmap Report, the linac would likely be constructed in two stages, with the second stage involving a major power upgrade. In the initial configuration, the linac could be built with the full complement of accelerating modules to reach 1-GeV output energy, but with only enough installed RF power to accelerate 1/4 of the nominal plant linac current, namely 11.25 mA. The accelerator tunnel would be constructed to the length needed for 1 GeV, and a periodic quadrupole focusing lattice would bridge the gap between the linac and the target high-energy beam transport system. The beam power is enough to drive several medium-power ATW demonstration transmuters or a single full-scale plant-size 840-MW,h, and would accommodate the relatively low transmuter reactivity levels that may be exist in early tests. After several years of operation of the DEMO facility with a single 840-MWth transmuter, the accelerator capability would be upgraded to 45 mA at 1 GeV, and up to three additional 840-MWth transmuters could be installed. RF beam splitters would be implemented to demonstrate delivery of beam power to separate transmuters. These steps would convert the DEMO facility into a prototype ATW plant, at 1/4 or 1/2 scale, The linac upgrade would involve adding RF power systems and a modification of the power distribution architecture. The two-stage approach to implementing the DEMO accelerator minimizes the capital outlay for the intial equipment installation, but also permits likely technology improvements and potential equipment unit-cost savings to be applied to the more powerful second stage.

3.1 Strawman Accelerator Design for ATW Plant The accelerators driving the ATW plants have a nominal beam power of 45 MW (45 mA CW, 1 GeV) each. There are two at each plant, with each one providing beam to four 840-MW~h subcritical assemblies. A strawman linac architecture is depicted in Fig. 1.3

35

ATW accelerator technology in US roadmap

above. As noted earlier, most of the accelerator would be superconducting, to achieve m i n i m u m capital and operating costs, and to minimize length. Accelerating gradients in the SC cavities have been assumed more than twice those in the A P T linac design, and the NC to SC transition has been pushed down to 11 MeV. A list of key parameters for the several sections of the superconducting linac is displayed in Table 3.1., The high energy portion of the SC linac (above 211 MeV) is made up of cryomodules containing elliptical 7 0 0 - M H z cavities and superconducting quadrupoles in a F O D O focussing lattice. There are two cavity beta values, 0.60 and 0.75. The low-energy portion of the SC linac is c o m p o s e d of cryomodules containing 3 5 0 - M H z spoke-type (1/2-wave) resonators and superconducting quadrupoles. There are three beta values, 0.20, 0.30, and 0.44. The unloaded cavity quality factor, Q0 is taken as 3x109 for all cavities. The N C front-end linac contains the proton injector, a 6.3-MeV 3 5 0 - M H z RFQ, and a short section of 350-MHz C C D T L (coupled-cavity drift-tube linac) structure to 1 1 MeV. Funneling arrangements at this energy could be a way to achieve independently adjustable current control at individual A T W transmuters, if needed. Simple on-off b e a m control to the individual targets could be obtained by a chopping system following the

~FQ. Table 3.1. Parameters for Strawman Superconducting A T W Linac

Cavity beta Input energy (MeV) Output energy (MeV) Energy gain (MeV) Cavity frequency (MHz) Cells per cavity Cavity length (m) Accelerating gradient (MV/m) Accelerating phase angle, phi (deg) Energy gain per cavity (MeV) RF power per cavity (MW) Aperture radius (cm) Evcak./Eacc Epeak (MV/m) Q~t Lattice period length (m) Cryomodule length (m) Cavities per cryomodule Klystron output power (MW) Number of klystrons Number of cryomodules Number of cavities

section 1 0.20 11.1 33.3 22.2 350 5 0.42 6.75 - 30 2.47 0.11 4 3.30 24.49 !.73E+06 10.9 9.8 9

section 2 section 3 section 4 section 5 0.30 0.44 0.60 0.75 33.3 77.8 21 I. 1 388.9 77.8 211.l 388.9 1100.0 44.4 133.3 ! 77.8 711.1 350 350 700 700 5 5 8 8 0.65 0.94 1.03 1.28 6.59 6.82 10.00 9.99 -30 -30 - 30 - 30 3.70 5.56 8,89 11.11 0.17 0.25 0.40 0.50 5 6 8 10 3.30 3.30 2.75 2.50 23.92 24.76 30.24 27.46 1.69E+06 1.75E+06 5.97E+05 5.96E+05 9.2 8. l 10.6 10.2 7.9 6.4 9.0 8.3 6 4 5 4

1

1

1

2

2

[

2

6

4

16

l

2

6

4

16

9

12

24

20

64

36

G. Lawrence et al.

Total length (m) Active fraction

11

18

48

43

163

0.35

0.42

0.47

0.48

0.51

Energy gain per meter (MeV/m)

2.0

2.4

2.8

4.2

4.4

Avg RF loss per section (W)

575

1125

3500

1528

6358

6

12

36

24

96

581

1137

3536

1552

6454

Static thermal load (W) Total refrigeration load (W)

3.1.1 Alternate Cryomodule Architectures Two approaches are being considered for the cryomodule architecture in the ATW high-energy linac. The one depicted in Fig.3.1 is aimed at minimum capital cost, featuring high power klystrons that each deliver RF power to four 8-cell 700-MHz cavities. This design is vulnerable from a beam reliability standpoint, since the fractional energy gain in each cryomodule is very large (44.4 MeV), and a beam abort would be required in the event of an RF station failure.

2.4-MWKlystronv 700MHz ~ . ~ ,,r,) Circulator Quadrupole

500kW~

I RFsplitter

500kW

500kW

I

500kW

Fig. 3.1. Cryomodule architecture for ATW plant linac emphasizing cost minimization.

The second design, illustrated in Fig.3.2 places the priority on beam reliability and puts cost in second place. In this cryomodule concept, short (4-cell) 700-MHz cavities are driven independently by medium power RF tubes, 300-kW high-efficiency IOTs similar to those used in the television industry. The energy gain in each cavity is 5.5 MeV, which is a low enough increment that it would not be necessary to interrupt the beam in the event of most RF station failures.

ATW accelerator technology in US roadmap

300-kW

,Otq,

37

Quadrupole Doublet

\

700-MHz 4-Cell Cavity

Fig.3.2. Cryomodule architecture for ATW plant linac that emphasizes beam reliability. This strawman linac concept takes full advantage of the advances in superconducting technology that have occurred in the past few years. It assumes thatthe high accelerating gradient levels recently reached at the frontier of SC cavity development will be standard for industrial fabrication by the time the DEMO linac is in preliminary design. In addition to the savings in electric power due to the elimination of resistive wall losses, SCRF cavity technology has other advantages for the ATW accelerator system. Specifically: •

•

•

•

• •

Higher accelerating gradients greatly reduces accelerator length, and number of components, providing cost and schedule benefits. The shorter accelerator also reduces other length-related costs such as that of the accelerator tunnel and klystron gallery. Larger bore radius in the accelerating structures becomes affordable, relaxing alignment, steering, and matching tolerances, and reducing beam loss and activation. This eases commissioning, fast recovery from faults and improves availability. Independently-phased short cavities have a large velocity acceptance, providing more flexible operation, allowing continued operation following component failure, and increasing overall system availability. It eliminates complex, high-throughput, high-maintenance water cooling systems, which are needed in a NC linac for removing the RF wall losses and also for cavity resonant-frequency control. The linac is built from a small number of identical components, providing cost and schedule benefits and also important maintenance benefits. A worldwide industrial capability now exists for fabrication of superconducting cavities and cryomodules. This benefits fabrication schedules.

Using spoke resonators of the type recently developed at ANL [7], most of the lowenergy portion of the linac (which in APT is made up of normal-conducting copper cavities) can also be superconducting, further reducing power losses and providing large apertures in this energy range. In APT beam simulations, this region has the highest probability of beam halo interception. Ultimately, it may be possible to push the NC-toSC linac transition energy down to the output energy of the RFQ, or perhaps even to build the RFQ itself as a superconducting device.

G. L a w r e n c e et al.

38

3.2 Beam Transport System. In the ATW plant, the beam is transported from the end of the linac to four subcritical fission transmuters. A possible beam transport layout is shown in Fig.3.3, which indicates two stages of beam splitting, providing four equal-current CW feeds to the transmuters. Cross-connects to the second set of transmuters fed by the second linac are not shown. For transmuter designs based on liquid targets and fuels, the beam entry would probably be from above, requiring a 90 ° vertical bending system in each transport channel; the transmuters would then be located several tens of meters below the accelerator beamline. Other transmuter designs, such as those using high-temperature gas-cooled technology, would permit horizontal beam entry, simplifying the beam transport arrangements.

location o f 90° vertical bends burner

beam

separation dipoles . o ~ J ~ r linac _...j - - -

..

separation dipoles

RF splitter i

:

I

-'-'-" c

~

.

RF splitter "

_ ~'~.----.--~ ~P-~-~----"-'~ =

~

)

~" -

.... n • • • "

I

P. "

r~_tuneup beam dump

I

10m

Fig.3.3. Schematic beam transport and distribution system to serve 4 transmuters.

4.0 Design and Trade Studies Needed for DEMO and ATW Plant Linacs There are a number of design trade studies that should be undertaken in the first two years of the ATW R&D program, to determine the optimum architecture for both the DEMO accelerator and for a full-size ATW plant. As indicated above, the DEMO accelerator would be built initially with sufficient power capacity to drive several medium-scale demonstration transmuters simultaneously, or a single full-size transmuter (840 MWth). In its final configuration, following the power upgrade, the DEMO accelerator would become a fully capable plant-scale unit, powering the ATW plant prototype with up to four 840-MWm transmuters. To the extent that it is practical, the accelerator upgrade would incorporate machine architecture and component improvements that have been realized during the period when the DEMO is in operation in its initial configuration.

ATW accelerator technology in US roadmap

39

The seven or eight full-size ATW plants that would eventually be constructed would each include two 45-mA, 1-GeV linacs, with a beam splitting arrangement to allow each linac to deliver equal-power beams to four transmuters. The transport system would be configured to allow each transmuter group to be supplied from either accelerator, insuring a high probability of continuity of electric-power generation. The kinds of pre-conceptual design and trade studies that need to be carried out would include: Development and analysis of strawman accelerator designs, including cryomodule architecture, RF system architecture, cryosystem concept, beam-sharing and control, and beam-transport architectures for ATW plants; • Analysis of optimum accelerator architecture to implement the DEMO stage at minimum cost, while accommodating a power upgrade for the prototype ATW plant. Assessment of basic accelerator parameters, such as cavity frequency, cavity gradient, cryogen temperature, RF generator size, RF coupler power, focusing lattice period and type, power supply size and configuration, etc.; • Beam dynamics analysis and simulations to assess optical matching requirements, beam halo minimization, and sensitivity to machine imperfections; • Mechanisms for assuring redundancy, invulnerability to component failures, and

rapid recovery from faults; • Fabrication and manufacturing strategies to reduce costs and improve component performance. Following the trade studies, a reference conceptual design will be produced for the upgradeable DEMO linac and beam transport system. This would be followed by preliminary and final design for the DEMO linac, and construction beginning in about FY09, according to the schedule in Fig. 1.3.

5.0 Accelerator E D & D Program for D E M O and A T W Plant

With construction of a DEMO facility beginning in 2009, there would be a total of 7 years to carry out the initial period of accelerator R&D supporting the DEMO linac design. Following construction of the DEMO linac in its initial power configuration, there would be an additional development period available from 2014 to 2021, in which the ATW plant accelerator design could be optimized. The power upgrade scheduled to begin in 2025 provides an opportunity for installing improved equipment that would be developed during this period.

40

G. L a w r e n c e et al.

Thus, the accelerator technology development program might be seen as divided into two time phases. The first, completed by late 2007, addresses components and subsystems that must be developed for implementation in the first phase of the DEMO facility. The second, would prepare a second generation of equipment to be incorporated in the highpower phase of the DEMO, when it is then to be converted to a prototype ATW plant. The total time spanning these two periods is so long that entirely new accelerator technologies may emerge that could significantly impact the accelerator reliability, performance, and cost relationship. 5.1 General Considerations

The APT accelerator ED&D (engineering development and demonstration) program provides a strong base for any future high-power linac design, including that for an ATW system, but has been aimed at supporting near term plant implementation of an APT plant (construction start in 2002). For this reason, component design choices for the APT linac were relatively conservative, and focused on moderate extensions of existing technology. Because of the longer time scale for implementation of the ATW DEMO facility, the supporting accelerator R&D program can and should look well beyond present component performance levels, and push in the directions that would provide a more optimized ATW plant design, one that has excellent performance, minimum cost, and high reliability. Therefore, the ATW accelerator R&D program would reach beyond what is being demonstrated in APT. The program should also should give priority to development of features needed for ATW systems that have not been a requirement in the APT plant, such as beam splitting and control hardware, and high beam reliability (as compared with high availability).

5.2 Key Technologies in A T W Accelerator ED&D Program

There are several accelerator technologies that need development and/or prototyping prior to final design of the DEMO linac. These include: • High-gradient (elliptical) high-beta SC cavities

• • • • • •

Low-beta (spoke type) SC cavities Short, large aperture SC quadrupoles Cryomodules for all sections of the SC linac, constructed using the above components SC beam splitters and septum magnets for distributing beams to separate transmuters Beam control using micropulse chopping systems, and/or funneling schemes A higher efficiency RF generator, the HOM lOT • Ultra-reliable key components, including RF power systems, injectors, magnets, DC power supplies, etc.

5.2.1 Superconducting RF Technology The ATW plants will need as high an accelerating gradient as practical for the superconducting cavities, and will also benefit from replacing most of the low-energy

ATW accelerator technology in US roadmap

41

normal-conducting accelerating structures in the APT linac design with superconducting cavities. Starting from APT accomplishments and the advancing edge of SCRF technology under development at several laboratories worldwide, the ATW DEMO schedule allows sufficient time to test high-gradient high-beta SC cavities and low-beta SC cavities for DEMO, as well as to prototype production cryomodules for all sections of the SC linac. SC quadrupoles similar to the RHIC (Relativistic Heavy Ion Collider) correction quadrupoles will be used in these cryomodules, and need to be prototyped.

5.2.2 RF Beam Splitter The most practical solution for distributing beam from the DEMO (and plant) accelerators to multiple targets is to use RF splitters that divide the beam intensity in two on a bunch-by-bunch basis. The splitters are several-meter-long SC cavities operating in a deflecting mode at an integer multiple of 1/2 the bunch frequency. During passage through these cavities, adjacent beam bunches are deflected by a few milliradians in opposite directions; after a short drift, the deflected beams enter septum magnets that complete the separation, and initiate the beam transport to the targets. For the ATW reference design, two sequential stages of RF splitters are needed, as shown in Fig. 1.1, to produce four beams of equal intensity.

5.2.3 Beam Interruptor It will be necessary to interrupt beam delivery to any plant transmuter, as needed to carry out refueling, fuel shuffling, or for maintenance. A crude method would be to simply deflect the beam away from the designated transmuter to a high-power beam dump. This would be costly both in terms of capital expense (for the high-power beam dumps) and wasted beam power. A more sophisticated approach, but one that needs considerable development, is to use a very high-speed beam deflector (chopper) following the RFQ. This device must be able to inhibit all the beam bunches destined for one or more of the four transmuters associated with each linac.

5.2.4 High-Efficiency RF Generator High-power CW klystrons in the 1-MW class are at present the standard RF power sources for driving high power linacs. Their effective DC-to-RF conversion efficiency is about 58% at full output power (allowing for control margin), the efficiency drops as the power demand is reduced, and they operate at high voltage (around 95 kV). A higher efficiency lower-voltage 1-MW RF source, the high-order-mode inductive-output tube (HOM IOT) is being developed as part of the APT ED&D program. This tube, if it performs as expected, will have 73% efficiency, the DC-RF efficiency remains high at reduced output, and the operating voltage is much lower (45 kV). This kind of tube would provide significant cost and performance advantages for an ATW linac and plant. Therefore, the HOM IOT testing and prototyping program should be pursued beyond the

42

G. L a w r e n c e et al.

APT ED&D program. The latter will develop only a first working HOM IOT version that will require significant improvement and long-term testing to become a viable replacement for the klystron.

5.2.5 High-Reliability Components In existing high-power accelerators such as the LANSCE linac, the beam is interrupted typically on the order of once per hour, due to RF-station faults, injector faults, or other equipment outages. Most of these interrupts are of short duration, less than a few seconds; a fraction of order 10% are of longer duration. Although the beam availability requirement for an ATW plant can be satisfied with an APT-type linac design, the beam continuity requirements are much more stringent, as discussed in section 1.2.1. A major objective of the ATW accelerator R&D program will be to understand the causes of beam interrupts in high-power linacs, and to reduce their frequency and duration to very low values. The need to minimize the number and duration of beam interrupts is a new requirement for accelerator design and operation, and has not previously had much emphasis. The development program to address this issue would require construction and long-term testing of representative high-reliability versions of RF power stations, injector, and other key components of an ATW linac. The LEDA facility at Los Alamos, built to demonstrate the APT linac front end would provide an excellent integrated test bed for these high-reliability components. This program would lead to equipment installed in the DEMO linac that has an extremely low fault rate in comparison with current experience. Operation of the DEMO facility itself will provide a long term evaluation of these designs and a rigorous testbed for further improvements.

5.2.6 Design for Lower-Cost Manufacturability Design for lower-cost manufacturability is a long-term effort that should be started in the initial ATW R&D phase, and continued beyond the first implementation phase of the DEMO linac. There are potentially large cost savings to be obtained in the deployment of the ATW plant linacs by a strong and sustained effort in this area. This program would be a combination of production-cost analysis, fabrication optimization, and manufacturing prototyping. The objective is to develop fabrication methods and processes for components employed in large quantities in a high-power accelerator (such as SC cavities, klystrons, power supplies, circulators, RF windows, vacuum elements, magnets, etc.) that should be significantly less costly than current approaches.

5.3 Legacy of APT Accelerator ED&D Program Since the accelerator proposed for the ATW Roadmap reference concept closely resembles the accelerator for the APT project, the ATW accelerator design will benefit greatly from the ED&D program that has supported the APT linac design, which has reached a high level of maturity. With the December 1998 DOE technology decision on

ATW accelerator technology in US roadmap

43

tritium production, the APT plant will not be built, but the project is placed in backup status and will complete the key elements of the ED&D program. Within the expected FY99 - 02 funding profile, the accelerator part of that program will: Demonstrate full-power operation of the Low Energy Demonstration Accelerator (LEDA), shown in Fig.5.1. LEDA is a complete prototypical front end of the 100-mA APT low-energy normal-conducting linac, including: - operation of a 6.7-MeV 350-MHz CW RFQ at 100 mA CW - operation of a short section of a 700-MHz CW CCDTL (8 MeV) at 100 mA CW confirmation of the match between these two structures - benchmarking of beam-dynamics design codes - beam halo measurements to confirm halo production model and simulations - operation of 1-MW CW RF power sources, power supplies, and RF distribution and control systems at 350 MHz and 700 MHz - beam diagnostics for measuring high-power beam properties at low energies - integrated system operation, control, and fault recovery schemes -

Demonstrate a complete prototypical medium-beta (~ = 0.64) cryomodule, one of the accelerating units in the high-energy APT superconducting linac, including: - performance of the 5-cell 13-- 0.64 superconducting cavities (5 MV/m at Q > 3x109) - performance of the high-power (adjustable) RF couplers (250 kW at 700 MHz) cryogenic performance (heat loads) at cavity operating temperature (2.15K) -

In addition, the APT ED&D program includes an initial demonstration of a 1-MW highefficiency RF generator (HOM lOT) that could eventually replace the klystron if successful. The operation of other RF power system components at high power levels will be also be confirmed, including RF vacuum windows, circulators, RF switches, vacuum/RF-waveguide valves, high-voltage power supplies, etc.

Fig.5.1 Low-energy demonstration accelerator (LEDA) shown with 6.7-MeV RFQ + injector.

44

G. L a w r e n c e et al.

6.0 Technology Development for ATW Accelerator 6.1 Injector and RFQ The LEDA facility at Los Alamos has recently demonstrated beam performance at 100 mA CW out of the 6.7-MeV RFQ [13]. Preliminary measurements show that the beam quality is within the range needed for the APT accelerator. This beam is more than adequate for injection into an ATW linac, so there is no ATW requirement for additional R&D in this area. Continuing APT-funded work is in progress to add a short section of the next NC accelerating structure, an 8-MeV CCDTL (Coupled-Cavity Drift-Tube Linac) to LEDA, and to carry out beam halo measurements. These activities will provide vital information for evaluation of an ATW linac design

6.2 SCRF Accelerating System 6.2.1 Overall Development Objectives The major objectives for an SCRF technology development program supporting the ATW DEMO linac and plant linac designs would be to: •

• • •

•

Reach much higher accelerating gradients in elliptical SC cavities (10 MV/m) than in the APT design. Such gradients have been considerably exceeded for beta =1 cavity designs, but have not been demonstrated for the beta < 1 elliptical multicell cavities needed in an ATW linac. Provide higher RF coupler power levels (500 kW/coupler) than in the APT design. Develop spoke-type SC cavities for the low beta region to increase the power efficiency in the linac from 11 - 211 MeV, and also to increase the beam aperture. Develop SC quadrupoles for use inside both the low-beta and high-beta cryomodules, to increase the transverse focusing strength and further increase overall linac power efficiency. Build and test prototypes of high-beta and low-beta high-gradient cryomodules.

6.2.2 SCRF Cavity Technology The use of SCRF technology based on niobium accelerating cavities operating at LHe temperatures (2.1K to 4.2K) reduces RF power loss by factors of typically 105 in comparison with normal-conducting copper cavity technology. This basic fact opens a large range of accelerator options and considerably broadens the design space. Most applications of SCRF technology to date have been for electron accelerators, such as the machines at Jefferson Laboratory (CEBAF), CERN (LEP II), KEK, and DESY. The

ATW accelerator technology in US roadmap

45

cavities for these electron accelerators, which operate at beta = l, have a basic elliptical shape that eliminates electron multipacting. The APT accelerator design uses similar elliptical cavities in the high-energy portion of the linac, but these are longitudinally compressed to compensate for the lower beta. ATW linac cavities would have similar shapes. The following discussion outlines the potential application of SCRF to an ATW linac.

6.2.2.1 Background SCRF technology is a mature, but still developing discipline. Particle beams were first accelerated with SC cavities nearly 35 years ago, with the first SCRF-based electron linac beginning operation more than 25 years ago, and the first SCRF ion linac over 20 years ago. In the past two decades, SCRF cavities have undergone vigorous and continuing development and are being employed in an ever-widening range of accelerator applications. From the beginning, two key characteristics exhibited by niobium SC cavity arrays have been: Excellent field performance. The present performance frontier is typically at peak surface (electric) fields in the range 15 - 20 MV/m, which provides accelerating fields of 7 - 10 MV/m, limited by field-emission electron loading. This is the operational state-of-the-art routinely achieved for a wide range of cavity geometries and frequencies, and employed in a number of operating electron and ion accelerators. •

High reliability. SCRF accelerators have proven to have generally high availability, and unscheduled maintenance has been low [14].

Recent advances in materials and cleaning techniques have led to demonstrations at several laboratories of peak surface electric field levels more than a factor of two higher than the above values [ 15,16]. These advances are summarized in Fig.6.1, which compares 1300-MHz beta -- 1 cavity test results obtained recently at TTF (Tesla Test Facility at DESY) with the installed cavity peak fields at 1500 MHz achieved in the CEBAF electron linacs (Jefferson Laboratory).

46

G. Lawrence et al.

Number CEBAF

'° f 6O

40

L TESLA, 1997 and 1998

\

2O 10 ~"

0

10

20

30

40

50

60

70

Epeak (MV/m)

Fig.6.1. Comparison of CEBAF and recent TTF cavity peak surface field performance.

In terms of medium-beta cavity tests, CEA/Saclay has recently achieved results with 700 MHz beta 0.65 single SC cells showing performance well above the ATW linac design objectives, as shown in Fig.6.2 below.

~ Proton s uperconductingCavity at cEA/s aclay.... i ::::-.

IE+12

F = 700MHz, 13 = 0.65

-. =T=4.2K" I ] OT=2.0K

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

i

1E+1t

-~- _I'e!,e-~!-!.~-:::-!!

i:-:.~

;

~!::!!!!:_:-:-!!! I ) T = I . 6 K

1

1E+10

!!!:! 1 1 E+09

?*~"~:=k _~!P:~i~_)/A~li!:::

i 1 E+08

ill i

0

i

i

--::ill

--::i ! _--:lii!! ~-i!! ]-_-!

i ill illl iii iiii ill L

i

5

i

t

10

i

i

i

i

!!

]

il iii:___~Qu~,a,]: i

15

Eacc (M e V / m )

i

i

20

i

i

i

i

25

L

i

i

30

ws~r~:~ r ~ e ~ : - ~ ~

Fig.6.2. Test results for Saclay beta = 0.65 single-cell 700-MHz cavity.

......

ATW accelerator technology in US roadmap

47

6.2.2.2 Advantages for ATW Application Compared to NC copper linac technology, SCRF linacs have significant advantages for achieving ATW linac performance objectives. SCRF linacs can have much higher accelerating gradients than NC linacs, because their RF wall losses are very small; with the high RF losses in copper cavities at room temperature, NC linac gradients (for CW operation) are restricted to relatively low values < 2 MV/m. High cavity gradients permit a shorter linac, which leads to a reduction in cost of the linac and in the cost of other length-related equipment and infrastructure. SCRF linacs have high power efficiency because of their very low RF losses. The amount of RF power and number of RF sources needed for the linac is significantly reduced in comparison with the equivalent NC linac. Even after adjusting for the cost and power requirements of the cryoplant needed to refrigerate the LHe coolant, an SCRF CW linac saves both in capital and operating costs. SCRF linacs accelerate beams using short independently-phased accelerating cavities that have a large velocity acceptance. This feature allows a superconducting linac to accelerate over a wide range of beam-energy profile along the linac. It offers flexibility of operation, making the linac tolerant to cavity and RF station failures, providing higher reliability and availability. SCRF linacs can have very large beam apertures, two-to-four times larger than equivalent copper structures. Large beam apertures practically eliminate beam activation of linacs due to halo beam losses, which assures hands-on maintenance of the accelerator.

6.2.3 SCRF Development Needs

6.2.3.1 APT Cryomodule Tests While there is a large experience base for beta = 1 SC cavities, relatively little development work has been done on elliptical cavities suitable for accelerating protons in the velocity range above beta = 0.5 and below beta = 1. The APT program is building a beta 0.64 cryomodule containing two 5-cell 700-MHz cavities (for acceleration in the energy span 211 - 471 MeV), and several cavities at different beta values have been built and tested at Saclay and KEK. However, RF power tests of a complete cryomodule are yet to be performed, and it is important for the ATW accelerator R&D program that such tests should be carried out, to confirm the integrated performance of such a system. Important data are needed in terms of attainable accelerating gradients and cavity Q values, power coupler performance, and the thermal loads on the refrigeration system. Figure 6.3 shows a picture of a 5-cell cavity for the APT beta 0.64 cryomodule.

48

G. L a w r e n c e et al.

I'l

ii

*

! _!i!:iiiii!ii i!iiii! ii

ii! iill!!!

Fig.6.3. Elliptical 5-cell beta = 0.64 cavity for APT prototype cryomodule, built by CERCA.

6.2.3.2. Power Couplers and RF Windows Because of the high beam current and high accelerating gradients in the ATW linac, it may be desirable to develop higher-power RF couplers. In one of the two strawman ATW cryomodule architectures discussed in Section 3.1.1, each cavity would have eight cells, requiring a coupler power of 500 kW, more than twice the rating of the APT power couplers. The APT cryomodule design has the complication of two power couplers for each cavity to provide the needed beam power, which leads to extra costs and complexity in the RF distribution system and also in the low-level RF controls. A power coupler of 500 kW or more transmitting capability would support the design of an ATW linac with higher accelerating gradients and simpler RF architecture. At the beginning of the APT accelerator design, power couplers had achieved 125 kW CW with beam. Recently, 500MHz power couplers at the KEK B-Factory have operated at up to 350 kW with beam, setting a new technology standard in this area. Selection of 500 kW as the design goal therefore seems a reasonable objective, and one that would pay dividends in reducing RF system complexity, improving reliability, and in reducing costs. In the APT accelerator design, the lifetime of the RF vacuum windows has a significant impact on estimated availability. Based on the limited database available for high-power windows, the relatively low lifetime of 20,000 hours has been assumed; this leads to a large number of window replacements operations each year, placing a heavy burden on accelerator maintenance resources and schedule. To address this issue, it is necessary to develop windows for the ATW linac that transmit high RF power and have estimated lifetimes in the range of several hundred thousand hours or higher.

ATW accelerator technology in US roadmap

49

6.2.3.3 SC Cavities for beta < 0.5

The use of SCRF technology in the velocity region beta < 0.5 would be advantageous if suitable cavities can be developed. A promising structure is the spoke-loaded cavity, several prototypes of which have shown promising results, but which needs to be developed in a configuration suitable for the frequencies and beam power required for an ATW linac. Figure 6.4 depicts a 350 MHz spoke cavity being developed at ANL which is suitable for 0.35 < beta < 0.6, and which has recently been tested at surface fields > 20 MV/m and accelerating gradients > 5 MV/m. To adapt this kind of cavity for ATW use, a high-power RF coupler would need to be developed, and other issues would have to be addressed.

Fig.6.4. Spoke cavity for 350-MHz application at ANL.

6.2.4 APT Linac SCRF Development Program 6.2.4.1 Beta = 0.64 cavity and cryomodule for APT

One of the major elements of the APT linac ED&D program is to build and and test a beta = 0.64 cryomodule containing two 5-cell elliptical cavities, operating at close to 5 MV/m accelerating gradient. This cryomodule, illustrated in Fig.6.5, will be the first unit developed specifically for use in a high-current proton linac. The goal of the ED&D task

50

G. Lawrence et al.

is to verify fabricability, performance including that of the cavities and couplers, and confirm the cryogenic loads. The cryomodule will serve as an integrated testbed of design concepts.

TUNEF

VESSEL SECTION OD = 1118 rnm L E N = 1575 m m

=WAVE GUIDE

W I D T H = 2~ I N N E R C OI

1=3~0mm

ivO

Fig.6.5. APT beta = 0.64 cryomodule design

The beta = 0.64 cryomodule design will be completed in FY 99, and it will be fabricated in FY 00. Starting in FY01, the cryomodule will undergo a series of tests to measure performance over a range of normal operating conditions and off-normal conditions. These tests are estimated to take one year. During the test program, design modifications will likely be made to improve performance. The ultimate objective is to test the operation of the cryomodule with full power RF; the level of completeness of this test program depends somewhat on the funding available as the APT project winds down.

6.2.4.2 Further SC cavity and cryomodule development 1. beta = 0.82 elliptical cavities and cryomodules If the APT program were to be restarted, 5-cell elliptical cavities at 13= 0.82 would be fabricated and tested, and then a four-cavity prototype cryomodule. Together with the performance data obtained for [3= 0.64 cavities and the beta = 0.64 prototype cryomodule, the information obtained would cover the complete range of beta values of

ATW accelerator technology in US roadmap

51

interest for the high-energy part of high-current proton linacs. In an ATW linac R&D program, a completed APT beta = 0.64 cryomodule could be used to test higher-gradient cavities and higher power couplers. For the ATW linac, the number of cells per cavity could be higher than in APT, and the coupler power could also be higher. Furthermore, the focusing elements may be inside the cryomodule, rather than in the inter-module warm regions, as in APT. Therefore, ATW linac cromodules will likely be significantly different in design and performance from APT linac cryomodules. To support design of an ATW DEMO linac, prototype ATW cryomodules would need to be built and tested at the nominal beta values of 0.60 and 0.75.

2. Spoke-loaded cavities and cryomodules for beta = 0.15 to 0.50 Superconducting niobium spoke-loaded cavities (1/2-wave coaxial-line resonators) of length and aperture suitable for the ATW low-energy linac need to be prototyped and tested over the range of beta values 0.15 to 0.50. The cavity designs needs to be correlated with beam simulation studies, and the trade off between aperture and cavity gradient optimized. Spoke cavities currently appear to be the best SC structure candidates for the low energy region of the ATW linac, but possibly other types of drifttube-loaded structures need to be evaluated and prototyped where indicated. The cryostats, RF couplers, cavity geometry, and operating temperature for the low beta linac sections will differ substantially from those in the elliptical cell cavity sections. For these reasons, it is desirable to design, develop, and test several complete cryomodules in this region. Because of their low design beta, beam tests with some of these cryomodules using the LEDA linac would be feasible, and should be given substantial priority.

3. High cavity gradients for elliptical cavities The ATW R&D program supporting the high-energy linac will build and test a number of single-cell 700-MHz elliptical cavities to demonstrate reliable and reproducible high field gradients (10 MV/m), for the nominal cavity design beta values (0.75, 0.60). This work will involve application of new niobium surface processing techniques. The techniques include high-pressure rinsing, electropolishing, heat treatment, and alternate chemical polishing approaches.

6.3. Superconducting Quadrupoles for ATW Cryomodules The quadrupole magnets needed for transverse beam control in the ATW linac will be similar to or identical to those needed for an earlier design of the APT superconducting linac. SC quadrupoles were designed for APT based on the successful design used in the Relativistic Heavy Ion Collider (RHIC) at BNL.

52

G. Lawrence et al.

The quadrupole R&D program for ATW would be very similar to the program originally planned for APT, and would be divided into two phases. In the first phase, 4 - 6 individual SC coils will be made and quench-tested. Four of these coils will then be assembled into a single prototype SC quadrupole and tested. In the second phase, as many quadrupoles as are needed for ATW cryomodule prototypes, plus a spare of each type, will be built and tested. To be properly phased with assembly of the first prototype cryomodules, this work should begin early in the 5-year R&D program.

6.3.1 Quadrupole Development The quadrupoles for the two different high-beta cryomodules in the ATW strawman have essentially the same design, with primarily dimensional differences, as shown in Table 2. The magnet is coil dominated, With the iron present to shield the field from the SC cavities. Figure 6.6 shows a cross section. The SC coils have a racetrack geometry, with an individual coil consisting of alternate layers of insulated wire and epoxy-impregnated fiberglass. The coils are conservatively designed to operate at half the maximum currentcarrying capacity of the wire. The largest contributor to field error is likely to come from the magnet-to-magnet variation of the integral gradient, G,L. Alignment of SC quadrupoles can be made as precise as that of resistive quadrupoles by utilizing fiducials on the cryomodules. The stress in the NbTi wire due to Lorentz forces is less than that in other magnets of similar construction. These magnets should have a very high mean time between failures, based on operational experience of the Tevatron proton-proton collider at FNAL, and the HERA electron-proton collider at DESY. Table 2. Superconducting Quadrupole Parameters Parameter

Medium-B

High-13

Aperture radius (cm)

6.5

8.0

Lattice half period (m)

1.70

2.03

Quadrupole length (m)

0.305

0.459

Quadrupole average gradient (T/m)

6.4 - 8.1

5.4

Nominal operating temperature (K)

4.7 - 5.3

4.7 - 5.3

ATW accelerator technology in US roadmap

53

APT SUPERCONDUCTING HIGH BETA QUAD COiL

.

.

.

.

.

.

COIL INSULATOR

BEAM

YOKE

TUBE

STA

:~ ¢42.0

-

- #16.0

BEAM TtJSE I.D,

YOKE O.D. KAWASAKI KHMNSOL STAINLESS STEEL COIL SUPPORT

PHOSPHOR BRONZE WEDGE (5"~

~LL B4~IE~ISIONS IN CNI

Fig.6.6. Cross section of the high-beta SC quadrupole cold mass.

6.4 RF Power Systems R&D

6.4.1 Introduction The RF power system considerations that went into the APT linac design [17] are all relevant to ATW requirements. This design produced an RF power system which efficiently employs the large amount of RF power required to produce a high-power CW beam at 1-GeV energy. As such it represents an excellent starting point for evaluating the RF system architecture needed for an ATW linac. Figure 6.7 sketches the RF system configuration employed in the APT high beta SC linac section. The ongoing APT program ED&D effort, including the completion of LEDA, the beta 0.64 cryomodule power tests, and RF component development and testing will form the basis for the R&D program supporting RF system development for the ATW DEMO linac and plant linac designs. This program should include: • Development and optimization of klystron HVDC power supplies using IGBTs (isolated-gate bipolar transistors), for improved reliability.

54

G. L a w r e n c e et al.

Reducing costs of high-power circulators, klystrons and other major RF system components, in cooperation with industry. • Development and testing of low-level RF (control) components. • Demonstration of a higher efficiency 300 kW to 1-MW CW RF generator (HOMIOT), followed extensive life test data on a successful design. • Extensive long-term lifetime testing of high-power RF windows. O

1-MW 700-MHz I~l~,l~t r t~r~

Resistive i n~cl

3-dB Hybrid ~nlitt~r

60 ° High-Power Ph~

~hifhar

Waveguide Switches /Yt P n r t ~

3-dB Hybrid .~nlittnr

RF Windows

Power Couplers 140-210 kW

Fig.6.7 APT 1-MW RF power module in high-beta superconducting linac

RF components would be tested through operation on LEDA and/or in dedicated RF test stands. An important issue is reliability, which can be greatly improved by elimination of spurious faults, and keeping system DC voltages as low as practical. Concurrent with the component R&D, an ongoing RF system architectural evaluation should take place as part of the accelerator pre-conceptual design and trade studies. It is expected that the bulk of the ATW accelerator RF system components would be constructed by industry. However, US industry lacks experience in many relevant areas, including high-power CW RF tubes, circulators, loads, and other RF components. Experienced construction of these components exists mainly in Europe and Japan. Thus

ATW accelerator technology in US roadmap

55

the development of US industry sources of ATW linac components could be a priority for supporting deployment a comprehensive ATW system. Development contracts should be awarded early in the R&D program.

6.4.2 Klystrons and Alternative RF Power Sources The necessary high-power klystrons (1-MW CW) in the ATW linac frequency range (350 to 700 MHz) have been built and demonstrated by industrial companies both domestic and foreign, although foreign experience in this area is significantly greater. In the time scale of the ATW plant linac deployment, it could be desirable to pursue development of higher power CW klystrons (2-MW or greater) which would provide a lower cost ATW linac design, and lead to a reduction in the number of RF system components. To support these higher power RF sources, there would be a concurrent need for higher power HVDC power supplies, circulators, etc. Design for lower-cost manufacturing could be pursued with great profit in this area, since RF system components are not presently built using mass-production assembly-line methods, and are very labor intensive. Given the 7 to 8 years of R&D that supports the initial incarnation of DEMO accelerator, followed by another 13 years before the RF power upgrade that converts the DEMO into a 1/2-scale ATW plant, there is an attractive alternative to the conventional klystron as an RF power source. That is the High-Order-Mode Inductive-Output Tube (HOM IOT). The IOT is already the preferred transmitter tube of the television industry, but at much lower output power levels (50 to 100 kW) than required for ATW. If a 300 kW-to-1 MW version of this tube can be developed it would have attractive possibilities for ATW service. It is more efficient, more compact, and less demanding than a klystron on its power supply. A program is already funded within APT (with a US industrial partner) to produce a first working HOM IOT prototype. The ATW R&D program should consider continuing this program and, after a successful tube design is achieved, proceed with extensive life testing and reliability engineering. Beam reliability issues enter strongly into the selection of an accelerator architecture and therefore the RF power system architecture. The RF system is a major cost driver, so cost minimization efforts would tend to push the ATW accelerator design in the direction of larger RF power components. However, as suggested in section 3.1.1, this may have a negative impact on beam continuity. Lower power components will likely provide an architecture with superior beam continuity, but will lead to higher costs. Thus, the issue of what size RF generators (and other components) should be developed for an ATW linac is an area that should be preceded by initial cost/reliability trade studies.

6.4.3 HVDC Power Supplies The selection of switch-mode power supplies based on IGBTs (Isolated Gate Bipolar Transistor) as the source of high-voltage DC power for the klystrons has been judged to be sound in technical reviews of the APT program. It is likely to lead to lower costs for

56

G. L a w r e n c e et al.

the prime power supply and to reduce the space requirements and infrastructure to house that power supply. The IGBT technology has been in use extensively at lower voltages, particularly for electric locomotives. Now these transistors are being produced at higher voltages and lower prices, and there is no engineering reason against stacking them together to reach the necessary voltages for powering the klystron electron beams. The APT ED&D program at Los Alamos and development programs at other labs are producing excellent results. The IGBT switch-based power supplies would be preferred for an ATW accelerator RF system, and luther power-supply developments using this technology are possible that could lead to RF system cost reductions and higher reliability.

6.4.4 RF Circulators High-power Y- junction type ferrite circulators at 350 MHz have been built for the APT LEDA project and are rated to operate into a fully reflective termination under the worst case scenario, when the klystron is transmitting full power (1.2 MW CW). A 700-MHz circulator of similar design is being developed by industry to operate at 1-MW CW power levels. The long-term performance and reliability of these devices should be part of an ATW RF system R&D program.

6.4.5 RF V a c u u m Windows RF vacuum windows made from alumina have been developed for APT based on the coaxial disk designs used in klystron service. Extensive high-power window testing programs at up to 1 MW CW transmitted RF power are underway at LANL as part of the APT ED&D program. The ATW R&D program would continue this testing in combination with the prototyping of ATW linac cryomodules to evaluate their expected performance in accelerator service. Also, alternative design and fabrication approaches for this key RF system component, such as rectangular waveguide windows would be investigated further. Some work has been initiated within the APT program. All in all, the RF power system components for an ATW accelerator are essentially in existence, developed for APT or previous accelerator applications, such as high-energy colliding beam machines. The R&D program should be continued in support of an ATW linac RF system design, but should focus on reliability, efficiency, and design for lowercost manufacture.

6.5 Beam Distribution and Control to Multiple Targets To address the requirement of serving multiple targets, the ATW accelerator R&D program should prototype a superconducting RF cavity beam splitter system, followed by an appropriate separation magnet and beam transport. RF splitter frequency, field

ATW accelerator technology in US roadmap

57

strength, cavity length and type are the principal issues. The beam splitting concept is sketched below in Fig.6.8.

RFDeflectCavi or tiesSeparatDiionpole ~ , ~

~ ~__~~__~

Fig.6.8. Left: Beam splitting scheme overview. Right: Bunch deflection in splitter cavity.

6.6 R&D and Accelerator Design to Address Beam Reliability

6.6.1 Accelerator Reliability and Beam Interrupt Tolerance For an A T W Plant, there is a premium on avoiding beam interrupts, especially those longer than a few seconds. This is a much stronger requirement than exists for the APT Plant accelerator, where the requirement is only that the annual beam availability should be large, say > 85%. For an ATW system there are two areas of concern. One has to do with minimizing the number and magnitude of transients on the power grid. When the beam trips off, there is a loss of more than 2000 MWe to the grid, which is not a small perturbation. The other concern is the step change in transmuter assembly temperature that occurs when the beam goes off and the fission power chain stops, and then again when the beam comes back on. After consulting with the SSI TWG, the following information has been developed. The time constants of concern for producing significant temperature excursions in the transmuter appear typically to be on the order of 1 second, and for the steam system driving the electric power generators to be on the order of several minutes. These time constants lead to the following restrictions on the allowed lengths of beam interrupts: • Beam interrupts of less than 0.1 second will have no effect at all on the electric power plant and will cause insignificant stresses in the transmuter assemblies. • Beam interrupts of from one second to a few seconds have no effect on the electric power plant, and will cause only small thermal shocks to the fuel assemblies. • Beam interrupts of tens of seconds will cause large temperature swings in the transmuter assemblies, but no effect on the power plant. Beam interrupts of a few minutes will require a slow ramp up of the beam (few minutes) to avoid thermal stresses in the subcritical assembly and will require active control of either beam current or a poison control rod for several additional minutes to compensate thermo/structural feedbacks. The power plant output begins to be

58

G. L a w r e n c e et al.

affected. Beam interrupts of > 10 min. will require a slow and laborious startup of the Rankine cycle electrical power plant which will take an hour or more. This will reduce the transmutation capacity factor very significantly if it occurs often. Great effort should be put into avoiding interruptions of this duration.

6.6.2 Beam Reliability Development Program The kind of reliability needed for an ATW accelerator must be addressed both through accelerator system design (tolerance to individual accelerating-unit failure, redundancy ot critical equipment, fast recovery from faults, etc.) and through equipment development (ultra-reliable components). Dedicated test stands should be built and operated, as appropriate, to evaluate the failure mechanisms for equipment that cause beam interrupts. Important elements include the proton injector, RF power systems and HVDC power supplies, accelerating cavities/cryomodules, focusing and transport magnets and power supplies, and diagnostics and control systems. These test stands will enable a detailed understanding of the failure mechanisms, will allow differentiation of real off-normal conditions from signal noise in protection circuits, and ultimately will allow elimination or compensation for the equipment failures. The goal is to dramatically decrease the frequency of beam interrupts in the ATW accelerator, by orders of magnitude from current experience. Some of the test stands will be attached to or embedded in LEDA and other APT-built accelerator facilities to provide integrated test-beds for ATW reliability testing and demonstrations. In the final analysis, operation and improvement of the ATW DEMO linac for many years will provide the ultimate tool for reaching the accelerator reliability goals that are needed in the ATW plants.

7.0 Program Costs, Schedule, and Deliverables 7.1 Program Costs The estimated costs of the 5-Year accelerator R&D and pre-conceptual design programs are shown in Table 1.1. Year 1 is the first year of the program.

ATW accelerator technology in US roadmap

59

Table 7.1. Cost (in $Millions) of the Accelerator 5-Year R&D and Pre-conceptual Design Programs. Program Element Reliability Analysis Reliability Testing SCRF Cavities High-Efficiency RF Tube RF Beam Splitter Preconceptual Design Totals

Year 1 1.5

3.0 1.0 1.0

Year 2 1.5 3.0 4.0 1.5 2.0

6.5

12.0

Year 3 1.0 5.0 5.0 1.5 1.5 2.0 16.0

Year 4

Year 5

7.0 3.0 0.5 1.0 2.0 13.5

5.0 2.0 0.5 0.5 2.0 10.0

Total 4.0 20.0 17.0 5.0 6.0 6.0 58.0

7.1. Schedule for Accelerator R&D

Figure 7.1 displays the roadmap for accelerator RD&D that addresses key accelerator performance and cost issues, and that also supports the DEMO accelerator design and construction, assuming the overall schedule in Fig. 1.4. The unshaded activities in FY 99 through FY02 are those that will be completed within the currently planned APT ED&D program, and the shaded activities are those to be funded from an ATW RD&D program. The RD&D activities are grouped into 1) component reliability development and testing, 2) integrated accelerator system reliability analysis and testing (including activities to be carried out using LEDA), 3) superconducting cavity and cryomodule development, 4) RF power systems development, and 5) development of beam distribution andcontrol devices. As noted elsewhere, the DEMO linac itself, when it comes on line, will be an integrated test bed for development of high reliability accelerator components, subsystems, and systems.

60

G. L a w r e n c e et al. 99

00

,'=

01

02

I

!

03

04

i'

LEDA operation/reliability tests

05

06

O7

O8

09

10

11

12

13

14

I i

Beta 0.64 cryomodule (APT funded) SpOke & elliptical high-grad SC cavities SC quadrupoles

Spoke-cavity ctyornodules High-gradient elliptical cavity cryomodUles

Advanced RF tube (HOM lOT) Higher-power ktystron & lOTs HV power supplies, circulators, loads, etc. Low - level FIF controls

RF windows and power couplers SC beam splitters

Mieropulse chopper syslem RF power station/cavity reliability testbed '

Injector reltabllity test stand Component reliability tests on Demo linac

!

J !

I I J I

I

I

m

Beam diagnostics (profile)

l APT Pro~lram

I

Fig. 7.1. Roadmap for accelerator RD&D programs that address key performance and cost issues and support the DEMO implementation schedule. 7.2 Program Deliverables The deliverables for the five-year accelerator R&D program and pre-conceptual design are seen as follows.

Reliability analysis Reports describing results of reliability analysis, including assessment of causes of unreliability in existing accelerators, and providing guidance for architecture and equipment design in ATW plant and demonstration linacs. Accelerator design trade studies to identify high-reliability architecture(s).

Reliability testing Q

Demonstrations of key accelerator equipment operating with ultra-low failure rates. Studies of key equipment systems, analyzing failure mechanisms, and progress in modifying these systems to improve reliability Guidelines for design and construction of the next generation of key equipment systems for ultra-low failure rates Demonstration of fast recovery from equipment failures and fast accelerator retuning schemes to assure minimal-duration beam interrupts [ 18,19]

ATW accelerator technology in US roadmap

61

Superconducting RF Cavity Development • • •

Multicell cavities operating at gradients of greater then 10 MV/m with high Q values. RF power couplers operating at levels required in pre-conceptual linac design. Development and testing (on LEDA) of low-beta superconducting cavities.

High-Efficiency RF generator •

Prototype RF generator operating with high DC to RF conversion efficiency and high reliability, at power levels required in pre-conceptual linac design.

RF Beam Splitter • •

Prototype multicell-cavity superconducting-RF beam splitter operating at the field required for adequate beam deflection. Demonstration of beam deflection on LEDA or LANSCE, with zero beam loss.

Pre-conceptual Design • •

8.0

Potential

Report describing pre-conceptual design of accelerator and beam transport system for an ATW demonstration facility. Preliminary cost estimates and schedule for implementation of the demonstration linac and beam transport system.

for

Collaborations

While the most extensive development and prototyping programs in high-power proton linacs are at Los Alamos, several related programs are in progress or are planned at other laboratories around the world, providing the potential for beneficial international collaboration. These are summarized below.

Saclay (Front end NC-linac demo; SC cavities) - 95-keV 100-mA proton injector (in operation) - 5-MeV 352-MHz RFQ (beginning fabrication) - 10-MeV 352-MHz short-section DTL (planned) - SCRF: tests of beta = 0.6 single-cell 704 MHz cavities (> 25 MV/m at Q > 5x10 ~°)

JAERI (Front end NC-demo & SC cavities) built and tested 2-MeV 200-MHz RFQ (80-mA peak current at 10% duty factor) - 200-MHz DTL cold model; power tests of short section at 20% DF SCRF: tests of beta = 0.5 single-cell 600-MHz SC cavity

-

CERN/INFN (SC cavity program) -

built and successfully tested sputtered 352-MHz 5-cell cavity at beta = 0.80 developing sputtered 352-MHz single-cell cavities for beta = 0.5 to 0.8

62

G. L a w r e n c e et al.

At CEA/Saclay in France, a strong team is building a front-end high-current CW linac prototype with objectives similar to the APT/LEDA program, and is also developing medium-beta high-gradient SC cavities and a prototype cryomodule similar to the one being developed in the APT SCRF program. Collaborations in high-power linac technology and superconducting cavities have been in place between Saclay and Los Alamos for several years, with staff being exchanged between institutions for significant periods of time, as well as several technical workshops that have been held to exchange information and plans. The present Saclay focus is directed toward an ATW demonstration facility that might be built in collaboration with other countries in Europe, and is looking at a 1-GeV 40-mA linac that would initially drive a 100-MWth test subcritical assembly, and then a larger prototype transmuter. As in the US Roadmap DEMO reference linac, the CEA linac would be a NC/SC hybrid accelerator, using highgradient (10 MV/m) superconducting cavities for the high energy section. JAERI and KEK in Japan have recently joined together to pursue a multipurpose research facility. The linac needed for this facility is intended to serve both as a driver for transmutation technology demonstrations, and also to supply a spallation neutron source for materials research and higher-energy accelerators for particle and nuclear physics. The linac design is a 1.1-GeV NC/SC combination like the ATW DEMO reference design, with similar parameter choices. In the R&D program, which has been underway for some time, JAERI has built and tested a pulsed 2-MeV RFQ, and plans to build and test a short DTL section. Single-cell SC cavities have been built at beta 0.5 and 0.9 and fabrication of 5-cell versions at these beta values is in progress. High fields have been achieved in vertical tests of the single-cell cavities. Collaboration on high-power linac design and hardware development has existed for several years between Los Alamos and JAERI, at a relatively modest level. Collaborations are being developed within Europe to pursue ATW accelerator technology R&D as well as integrated demonstrations. These are being put together within an EC framework, and aimed at a staged ATW system demonstration similar to that discussed in the US Roadmap. Given the current activities and plans for accelerator design and development in the US, Europe, and Asia, there is an excellent opportunity for fruitful collaborations in this area, and possibly some division of effort in order to reduce costs.

9.0 Summary A Roadmap has been established for an accelerator R&D program that will support an ATW DEMO linac and plant linac designs. The APT accelerator design and ED&D program provide a strong technology base for an ATW accelerator. The ATW linac design, while starting from this platform, has additional and different performance challenges. In particular, the ATW beam must have very high reliability; with much fewer beam interrupts per year than in present accelerator operations. Also, the ATW linac output beam must be distributed to multiple transmuters (using RF splitters)

ATW accelerator technology in US roadmap

63

introducing new control requirements for the individually transported beams. A linac reference design has been proposed that aims at higher accelerating gradients and lower NC/SC transition energy (than in APT), to increase efficiency & reduce costs. A national/international program can develop ATW linac designs and carry out the accelerator R&D.

10.0 References

1. 2. 3.

4. 5.

6.

"A Roadmap for Developing Accelerator Transmutation of Waste (ATW) Technology," A Report to Congress, DOE/RW-0519, October 1999. "A Roadmap for Developing ATW Technology: Accelerator Technology," LA-UR99-3225; Appendix to Ref. 1. G.P. Lawrence, "High-Power Proton Linac for APT: Status of Design and Development," 1998 International Linear Accelerator Conf., Chicago, August 23-28, 1998, 26-30. D. Schneider, "A Review of High Beam Current RFQ Accelerators and Funnels," 1998 European Particle Accelerator Conf., Stockholm, June 22-26, 1998, 128-132. T.P.Wangler et al, "Beam Dynamics Design and Simulation Studies of the APT Superconducting Linac," 1999 Particle Accelerator Conf., New York, March 29April 2, 1999, 611-613. K.C.D.Chan et al., "Engineering Development of Superconducting RF Linac for High Power Applications," 1998 European Particle Accelerator Conf,, Stockholm, June 22-26, 1998, 1843-45.

7. 8.

K. Shepard et al., "SC Driver Linac for a Rare Isotope Facility," ibid. J.D.Schneider, "Initial Testing of the RFQ and Related Equipment," Proc. of OECD/NEA Workshop on Reliability and Utilization of High-Power Accelerators, Aix-en-Provence, France, November 22-24, 1999, to be published. 9. K.C.D.Chan et al., "Review of Superconducting RF Technology for High-Power Proton Linacs." Ninth Workshop on RF Superconductivity, Santa Fe, November 1-5, 1999, to be published. 10. Th.Stammbach, S.Adam, H.R.Fitze, W.Joho, and U.Schryber, "Potential of Cyclotron-Based Accelerators for Energy Production and Transmutation," International Conference on Accelerator-Driven Transmutation Technologies and Applications, Las Vegas NV, July 1994, LAUR-95-1792, p 229-235. 11. Th.Stammbach, S.Adam, T.Blumer, A.Mezger, P.A.Schmelzbach, P.Sigg (PSI), "The Cyclotron as a Possible Driver for ADS Systems," Proc. of OECD/NEA Workshop on Reliability and Utilization of High-Power Accelerators, Aix-enProvence, France, November 22-24, 1999, to be published. 12. Proceedings of Workshop on Critical Beam-Intensity Issues in Cyclotrons, December 3-6, 1995, Santa Fe, NM, Los Alamos Report LAUR-96-1492, 1996.

64

G. Lawrence et al.

13. J.D.Schneider, "Initial Testing of the RFQ and Related Equipment," Proc. of OECD/NEA Workshop on Reliability and Utilization of High-Power Accelerators, Aix-en-Provence, France, November 22-24, 1999, to be published. 14. C.Reese, "Overview of CEBAF Operations and SRF-Related Activities at Jefferson Lab," 1999 Workshop on RF Superconductivity, Santa Fe, November 1-5, 1999, to be published. 15. C.Z.Antoine, "Alternative Approaches for Nb Superconducting Cavities Surface Treatment," ibid. 16. L.Lilje et al., "Performance Limitations in SC Cavities at TTF-Current Studies and Future Perspectives," ibid. 17. M.T. Lynch, "RF System Developments for CW and/or Long Pulse Proton Linacs," 1998 International Linear Accelerator Conf., Chicago, August 23-28, 1998, 1021-25. 18. S.Simrock, "Advances in RF Control for High Gradients," 1999 Workshop on RF Superconductivity, Santa Fe, November 1-5, 1999, to be published. 19. H.Safa, "Reliability: A Challenge for Superconducting-Cavity Technology," Proc. of OECD/NEA Workshop on Reliability and Utilization of High-Power Accelerators, Aix-en-Provence, France, November 22-24, 1999, to be published.