ICANS-XVIII
18th Meeting of the International Collaboration on Advanced Neutron Sources
April 25-29, 2007
Dongguan, Guangdong, P R China
ACCELERATOR DESIGN FOR ADS
J-L. Biarrotte1
On behalf of the EUROTRANS WP1.3 collaboration
1
CNRS / IN2P3 / Institut de Physique Nucléaire d’Orsay, France
Abstract
An Accelerator Driven System (ADS) for transmutation of nuclear waste typically requires a
600 MeV - 1 GeV accelerator delivering a proton flux of a few mAs for demonstrators, and of a few tens of
mAs for large industrial systems. Such a machine belongs to the category of the high-power proton
accelerators, with an additional requirement for exceptional "reliability": because of the induced thermal
stress to the subcritical core, the number of unwanted "beam-trips" should not exceed a few per year, a
specification that is several orders of magnitude above usual performance. Consecutive to the work of the
European Technical Working Group (ETWG) on ADS, the Preliminary Design Study of an Experimental
ADS (PDS-XADS) was launched in 2001 as a 5th Framework Program EC project1. A special Working
Package was dedicated to the accelerator design, taking especially into account that the issue of
“beam-trips” could be a potential “show-stopper” for ADS technology in general. A reference solution,
based on a linear superconducting accelerator with its associated doubly achromatic beam line, has been
worked out up to some detail. For high reliability, the proposed design is intrinsically fault tolerant, relying
on highly modular “de-rated” components associated to fast digital feedback systems. This paper describes
such a reference design, and underlines the R&D to be performed to demonstrate the feasibility of the
proposed solution. This work is supported by the 6th Framework Program EC project "EUROTRANS"2.
1. Introduction
The basic purpose of Accelerator Driven Systems (ADS) is to reduce the nuclear wastes’ radio-toxicity,
volume and heat load before their underground storage in deep geological depositories. This issue is
particularly significant in Europe where 2500 tons of spent fuel are produced every year by the 145 reactors
of the European Union [1].
The proposed solution relies on the “partitioning” & “transmutation” strategy: the different elements of
the spent fuel are chemically separated, minor actinides, plutonium and long lived fission products are
isolated and recombined to obtain new fuel assemblies to be used and “burnt” in dedicated “transmuter”
systems. Such a strategy should significantly – by orders of magnitude – decrease the long term radio-toxicity
of the spent fuel: with a separation efficiency of 99.9% of the long-lived products from the waste, followed
by transmutation, the natural uranium ore radioactivity level can be reached in less than 1000 years, instead
of millions of years if no action is performed [1].
An ADS transmuter system is composed of two main parts: a sub-critical reactor (keff < 1), in which the
chain reaction can not be self-sustained, that greatly relieves the safety problem aspect, and an intense
spallation source that provides the “missing” neutrons needed to keep the reaction going on. Such a neutron
source, composed by a liquid lead target subjected to a high energy proton flux, also produces the suited
broad energy spectrum required to “burn” the minor actinides components, that are otherwise accumulated in
conventional thermal spectrum critical reactors.
1
2
EC Contract N° FIKW-CT-2001-00179, “PDS-XADS”
EC Contract N° FI6W-516520, “EUROTRANS”
336
2. The EUROTRANS research programme
The EUROpean research programme for the TRANSmutation of high-level nuclear waste in accelerator
driven systems (EUROTRANS) is funded by the European Commission within the 6th Framework
Programme, and involves 31 partners (research agencies and nuclear industries) with the contribution of 16
universities. EUROTRANS is a 4-year programme (2005-2009) extending previous activities (PDS-XADS,
Preliminary Design Study for an eXperimental Accelerator Driven System) and paving the road towards the
construction, during the next EC framework programmes, of an eXperimental facility demonstrating the
technical feasibility of Transmutation in an Accelerator Driven System (XT-ADS).
Within the EUROTRANS programme, the activities are split into five main technical areas (called
Domains), respectively devoted to: the design of the ADS system and its sub-components; small-scale
experiments on the coupling of an accelerator, a spallation target and a sub-critical core; studies on advanced
fuels for transmuters; investigations on suited structural materials and heavy liquid metal technology;
collection of nuclear data for transmutation. The main objective is to work towards a European
Transmutation Demonstration (ETD) in a step-wise manner, i.e.:
• provide an advanced design of all the components of an XT-ADS system, in order to allow its realisation
in a short-term view (~10 years);
• provide a generic conceptual design of modular European Facility for Industrial Transmutation (EFIT)
for the long-term objective of the program.
3. Machine design for the European Transmutation Demonstration
The XT-ADS machine, that will be loaded with conventional MOX fuel, is meant to be built and tested in
the near future (before 2020) so as to fulfil three main objectives: 1. demonstrate the ADS concept (coupling
of proton accelerator, spallation target and sub-critical assembly) at significant but reasonable core power
levels (50 to 100 MWth); 2. demonstrate the transmutation using some dedicated positions able to accept
minor actinides assemblies; 3. provide a multi-purpose irradiation facility for the neutron community in
general, and for the testing of different EFIT components in particular (samples, fuel pins…).
The EFIT facility will be an industrial-scale transmutation demonstrator system, loaded with
transmutation dedicated fuel. Its characteristics are meant to maximise the efficiency of transmutation, the
easiness of operation and maintenance, and the high level of availability in order to achieve an economical
transmutation. Despite these sometimes contradictory objectives, the XT-ADS and the EFIT machines share
the same fundamental system features. Especially, both designs use (cf. Table 1):
• a superconducting linac solution to produce the required high-power proton beam; the main reasons for
this choice are the power-upgrading capability of this solution, and the perspectives of improvement of the
beam reliability;
• a liquid metal core coolant and spallation target; the metal is pure lead for the EFIT design, and
Lead-Bismuth Eutectic (LBE) for the XT-ADS, allowing lower working temperatures.
4. The reliability-oriented reference accelerator for ADS
The European Transmutation Demonstration requires a high-power proton accelerator operating in CW
mode, ranging from 1.5 MW (XT-ADS operation) up to 16 MW (EFIT). The main beam specifications are
shown in Table 2. At first glance, the extremely high reliability requirement (beam trip number) can
immediately be identified as the main technological challenge to achieve.
The reference design for the accelerator has been developed during the PDS-XADS programme [2,3] and
is based on the use of a superconducting linac (see Figure 1). Such a choice allows to obtain a very modular
and upgradeable machine (same concept for prototype and industrial scale), an excellent potential for
reliability, and a high RF-to-beam efficiency thanks to superconductivity (optimized operation cost). For the
injector, an ECR source with a normal conducting RFQ is used up to 3 or 5 MeV, followed by an energy
booster section which uses normal conducting H-type DTL or/and superconducting CH-DTL structures up to
a transition energy still under optimisation, around 20 MeV. This first part of the linac is duplicated in order
to provide good reliability perspectives. Then a fully modular superconducting linac, based on different RF
structures (spoke, elliptical), accelerates the beam up to the final energy (350, 600, 1000 MeV…). Finally a
doubly-achromatic beam line with a redundant beam scanning system transports the beam up to the target.
337
Table 1: Baseline characteristics of the XT-ADS and EFIT machines.
GOALS
MAIN
FEATURES
XT-ADS (ADS Prototype)
EFIT (Industrial Transmuter)
Demonstrate the concept
Demonstrate the transmutation
Provide an irradiation facility
Maximise the transmutation efficiency
Easiness of operation & maintenance
High level of availability
50 – 100 MWth power
Several 100 MWth power
Keff around 0.95
Keff around 0.97
600 MeV, 2.5 mA proton beam
(back-up: 350 MeV, 5 mA)
800 MeV, 20 mA proton beam
Conventional MOX fuel
Minor Actinide fuel
Lead-Bismuth Eutectic coolant & target
Lead coolant & target (back-up: gaz)
Table 2: XT-ADS & EFIT proton beam general specifications.
XT-ADS
EFIT
Max. beam intensity
2.5 – 4 mA
20 mA
Proton energy
600 MeV
800 MeV
Vertical from above
Beam entry
Beam trip (>1sec) number
< 5 per 3-month operation cycle
< 3 per year
Beam stability
Energy: ±1%, Intensity: ±2%, Size: ±10%
Beam time structure
CW with 200μs low frequency 0-current holes
Figure 1: The reference accelerator scheme for the XT-ADS (& EFIT).
338
The ADS accelerator is expected – especially in the long term EFIT scenario – to have a very limited
number of unexpected beam interruptions per year, which would cause the absence of the beam on the
spallation target for times longer than one second. This requirement is motivated by the fact that frequently
repeated beam interruptions induce thermal stresses and fatigue on the reactor structures, the target or the fuel
elements, with possible significant damages, especially on the fuel claddings; moreover these beam
interruptions decrease the plant availability, implying plant shut-downs in most of the cases. Therefore, it has
been estimated that beam trips in excess of one second duration should not occur more frequently than five
times per 3-month operation period for the XT-ADS, and three times per year for the EFIT.
To reach such an ambitious goal, which is lower than the reliability experience of typical accelerator
based user facilities by many 2 or 3 orders of magnitude, it is clear that reliability-oriented design practices
need to be followed from the early stage of component design. In particular:
• “strong design” practices are needed: every linac main component has to be de-rated with respect to its
technological limitation (over-design);
• a rather high degree of redundancy needs to be planned in critical areas; this is especially true for the
identified “poor-reliability” components: linac injector, and RF power systems, where solid-state amplifiers
should be used as much as possible;
• fault-tolerance capabilities have to be introduced to the maximum extent: such a capability is expected
in the highly modular superconducting RF linac, from at least 20 MeV [9].
A preliminary bottom-up reliability analysis (Failure Mode and Effects Analysis, FMEA) has been
performed in order to identify the critical areas in the design in terms of impact on the overall reliability [11].
This activity confirms the choice to provide a second, redundant, proton injector stage (composed of the
source, RFQ and low-energy booster), with fast switching capabilities. After the injector stage, the
superconducting linac has a high degree of modularity, since the whole beam line is an array of nearly
identical “periods”. All components are operating well below any technological limitation in terms of
potential performances, and therefore a high degree of fault-tolerance with respect to cavity and magnets can
be expected in the superconducting linac, where neighbouring components have the potential to provide the
functions of a failing component without affecting the accelerator availability. Clearly this approach implies
a reliable and sophisticated machine control system, and in particular a digital RF control system to handle
the RF set points to perform fast beam recovery in the case of cavity failures.
5. R&D activities for ADS accelerator development
The EUROTRANS accelerator work-package (WP1.3) is split in several tasks, focused to design the
main subcomponents of the ADS linac in the different energy areas, and to identify their reliability
characteristics. Activities are dedicated to:
• experimental evaluation of the proton injector reliability, performed at the IPHI injector,
• assessment of the reliability performances of the intermediate-energy accelerator components, with
particular attention to the comparison of different accelerating structures,
• design and experimental qualification of the reliability performances of a high-energy cryomodule
tested at full power and nominal temperature,
• design and test of a prototypical RF control system intended to provide fault-tolerance operation of the
linear accelerator,
• update of the accelerator design, including beam dynamics issues and investigation of control strategies
for fault-tolerance, development of more reliability analyses and cost estimations for the XT-ADS and EFIT.
5.1 Proton injector reliability tests
The IPHI injector (see Figure 2), developed in France at Saclay by CEA and CNRS, fulfils the
specifications of the ADS proton injector, with wide margins in term of beam current capabilities. It is
composed by the 95 keV SILHI ECR ion source, an already tuned low-energy beam transport line, a 3 MeV
copper RFQ under final completion, and the associated diagnostic beam line [4].
In the past years, the SILHI source has been successfully used for several week-long reliability tests at
currents of 30 mA, showing no beam stops and occasional sparks in the extraction region, causing no beam
interruptions. In the EUROTRANS context, these tests will be extended, to include acceleration in the RFQ
339
and propagation in the beam lines: the IPHI injector, once completed in 2008, will be used for a long-run test
(2 months), to demonstrate and assess, on a real scale, the reliability characteristics of this accelerator
subcomponent.
Moreover, the possibility to achieve the sharp 200 μs “beam holes” at low repetition rate (10-3 to 1 Hz),
that is required for the sub-criticality monitoring of the ADS core, has been successfully tested directly
pulsing the SILHI source. Encouraging fall and rise time durations of less than 30 μs have been obtained.
This test will also be extended to the full beam of the IPHI injector, at 3 MeV.
Figure 2: Scheme of the IPHI injector, and picture of the diagnostic beam line installation in Saclay.
5.2 RF structures for the intermediate-energy section
For the intermediate-energy region after the RFQ and up to approximately 100 MeV, several cavity types
are considered as valid “candidates”. Studies and tests of prototypes are being developed in order to evaluate
their feasibility and assess their potential reliability performances. In particular, activities concentrate on: 1.
high shunt impedance copper DTL structures of the IH and CH type, using “Konus” focusing scheme
(focusing elements outside the drift tubes) that provides high real estate gradients; 2. superconducting
multi-gap CH structures, with “Konus” focusing scheme, ensuring both high real estate gradients and
optimized RF-to-beam efficiency thanks to superconductivity; 3. superconducting spoke cavities, which are
very modular, providing some fault-tolerance capability, and can operate efficiently from very low energies
(around 5 MeV) up to 100 MeV and more. The two first structures are meant to be part of the redundant
injector front end, while the spoke structures will be part of the fault-tolerant independently-phased linac. At
the present phase of the project, the transition energy is around 17 MeV, still to be optimised given the results
of the on-going R&D.
Figure 3: Pictures of the SC CH-cavity during vertical test at Frankfurt (left) and of the spoke horizontal
test cryomodule during its installation at Orsay (right).
340
The first activity is carried out by IBA and IAP Frankfurt and is still in the design phase. Concerning the
second activity, a 19-gap superconducting 350 MHz CH prototype has been successfully tested at IAP
Frankfurt in vertical test (Figure 3) and is ready to be installed in an existing horizontal cryostat, with an
external driven mechanical tuner for further testing [5]. Concerning the spoke activities, CNRS / IPN Orsay is
equipping an existing cryomodule (Figure 3) with a 350 MHz β=0.15 spoke cavity, fully dressed with its
stainless steel helium reservoir, its cold frequency tuning mechanism and its high power (20 kW) RF coupler
[6]. The cavity, which has already reached the design goals in vertical tests, once completed with its ancillary
RF components, will be operated in 2008 at nominal operating conditions in long tests to determine the
reliability characteristics of the components.
5.3 The high-energy cryomodule
The high-energy part of the linac uses low-beta elliptical cavities from an energy of approximately 100
MeV, using 3 different beta-sections (0.5, 0.65, 0.85). Such a technology is already successfully used
worldwide (e.g. at the SNS), but the full demonstration of the very low-beta section is not yet fully
accomplished. Excellent test results have been achieved with the β=0.47 TRASCO cavities at INFN Milano
[7], but besides the development of the bare superconducting cavity, it is important to prototype and test each
auxiliary system needed for the cavity operation in a real environment (power coupler, RF source, power
supply, RF control system, cryogenic system, cryostat…).
The goal of this EUROTRANS third task is thus to design, build and test before 2009 an operational
prototypical cryomodule of the high-energy first section of the proton linac [8]. This cryomodule will host
one of the existing β=0.47 TRASCO cavities, equipped with a cold frequency tuner, developed by INFN, and
a 150 kW CW RF coupler, developed by CNRS. This cryostat, under development at INFN and CNRS, will
be assembled and tested at IPN Orsay under nominal 2K operating conditions, and extensive tests (without
beam but at full 80kW RF power level) will be performed to qualify its reliability characteristics. Figure 4
shows a conceptual layout of the module and of the RF power coupler.
Figure 4: Conceptual layout of the single cavity module (left) and associated power coupler (right) for the
testing of high-energy components.
5.4 Low Level RF control system and fault-tolerance
The scope of this activity is to develop a digital Low Level RF control system intended to provide the
necessary field and phase stability (±0.5% in amplitude, ±0.5° in phase) and to identify and handle fast
recovery scenarios for cavity failures in the superconducting linac, by means of local compensation at
neighbouring cavities.
This local compensation method (cf. Figure 5), in which neighbouring components have the potential to
provide the functions of a failing component, has been demonstrated on the beam dynamics point of view
from 10 MeV [9], given that modular independently-phased accelerating cavities are used and that some RF
power margin (up to 30%) is available. Such a fault-recovery system has even been recently demonstrated on
341
real operation at the SNS for high-energy beams (>200MeV) [10], using a global compensation method
where all the linac components are retuned, but with the advantage of requiring very few margins in terms of
RF power. The remaining step is now to identify and develop fast failure recovery scenarios to ensure that
such a retuning is performed in less than 1 second; this would probably require: fast fault detection and beam
shut-down; fast communication between neighbouring LLRF systems; fast update and tracking of the new
field and phase set-points; adequate management of the tuner of the failed cavity; final beam recovery.
Digital techniques become necessary to meet the speed and software configuration required by such a
retuning procedure. A FPGA-based digital LLRF control system, in which a number of key functionalities
are implemented on a single chip, offers indeed a high-grade reliability and flexibility. Developments are
going on in CEA and CNRS at 704 MHz and 352 MHz respectively, with very encouraging preliminary
results [6]. Figure 5 shows a picture of a first prototype developed by CNRS (LPNHE/IPNO).
Figure 5: Principle of the local compensation method (left) and picture of a prototypical FPGA-based
digital LLRF control card developed at CNRS (right).
5.5 Accelerator design
The last task in the accelerator work-package deals with the progress of the accelerator design and the
characterization of its reliability characteristics, the final goal being to obtain in 2009 a frozen detailed
conceptual design of the XT-ADS linac, with assessed reliability and costing.
Start-to-end beam dynamics simulations are being performed, especially for the different options for the
intermediate-energy structures. The modelling also includes in the independently-phased linac the effect of
the beam transients induced by RF faults, and explores the possible recovery scenarios needed to provide the
necessary beam availability to the ADS application. The design of the final beam line connecting the linac to
the reactor is also on-going, based on the previous PDS-XADS design guidelines.
Furthermore, an integrated reliability analysis is being performed to predict the reliability characteristics
of the proposed accelerator, following the standard practices used in the nuclear energy applications (mainly
failure mode and effects analysis and reliability block diagrams), and integrating the experimental results and
demonstrated reliability numbers. This activity is the continuation of the PDS-XADS work, that already
showed that a reliability-oriented design can, without necessarily improving the “Mean Time Between
Failure” values (MTBF) of each sub-system, greatly improve the MTBF of the whole system by 1 or 2 orders
of magnitude, especially with the implementation of redundancy, fault-tolerance and corrective maintenance
strategies in critical areas [11,12].
The deduced and optimised design will then provide the basis for the cost estimation for XT-ADS and
EFIT and a possible schedule for its realisation. This costing activity has already been started with parametric
studies to test the influence of the final energy on the linac price, especially showing that a 600 MeV solution
appears to be only 20% more expensive that a 350 MeV solution.
6. Conclusion
This paper describes the ADS accelerator reference solution, based on a reliability-oriented linear
superconducting accelerator, and summarizes the R&D presently on-going on some prototypical accelerator
components. Compared to warm temperature linacs or circular machines, a superconducting linac, much
more “forgiving”, has a great potential for high reliability, that is the main requirement for the ADS
application.
The “less than a few beam trips per year” goal is still 2 or 3 orders of magnitude below present accelerator
performance, but seems to be reachable. As a matter of fact, the concept of fault-tolerance in a
342
superconducting linac with independently-phased cavities is now proven, at least at high energies, by the
SNS experience, where a beam recovery system is successfully running [10]. An integrated reliability
analysis of the whole system shows that this SNS proof of principle of the fault-tolerance concept potentially
improves the reliability figure by at least 1 order of magnitude already (given that we can implement it in less
than the 1 second limit) [12]. The remaining order of magnitude will be gained by redundancy, corrective
maintenance plans, careful design of auxiliary systems, and by improving the MTBF of individual
components using the results from the on-going R&D performed within EUROTRANS.
Aknowledgements
The principal institutions in the accelerator work-package WP1.3 of EUROTRANS are CNRS (F), CEA
(F), IBA (B); IAP-Frankfurt University (D) and INFN (I). Additional contributions, especially in issues
dealing with the interface accelerator / reactor, are provided from ANSALDO (I), AREVA NP (D), ITN (P)
and UPM (S). The program has the financial support of the European Commission through the contract
FI6W-CT-2004-516520.
References
[1] European Technical Working Group, “The European Roadmap for Developing ADS for Nuclear Waste
Incineration”, ISBN 88-8286-008-6, ENEA 2001.
[2] A.C. Mueller, “The PDS-XADS reference accelerator”, International Workshop on P&T and ADS
Development, Oct 2003, SCK•CEN, Belgium.
[3] J-L. Biarrotte et al., “A reference accelerator scheme for ADS applications”, Nuclear Instruments and
Methods A 562 (2006) 565-661.
[4] P.Y. Beauvais, “Recent Evolutions in the Design of the French High Intensity Proton Injector (IPHI)”,
Proceedings of the EPAC 2004 Conf., July 5-9, 2004, Luzern, Switzerland.
[5] H. Podlech, “Development of Room Temperature and Superconducting CH-Structures for High Power
Applications”, 5th OECD NEA International Workshop on Utilization and Reliability of HPPA, May
2007, Mol, Belgium.
[6] A. Ponton, “Development of Superconducting Spoke Cavities for an ADS Linac”, 5th OECD NEA
International Workshop on Utilization and Reliability of HPPA, May 2007, Mol, Belgium.
[7] N. Panzeri, “Status of the Preparation of the Elliptical Cavity System for the EUROTRANS Cryomodule”,
5th OECD NEA International Workshop on Utilization and Reliability of HPPA, May 2007, Mol,
Belgium.
[8] S. Barbanotti, “Design of a Test Cryomodule for the High-energy Section of the EUROTRANS Linac”,
5th OECD NEA International Workshop on Utilization and Reliability of HPPA, May 2007, Mol,
Belgium.
[9] J-L. Biarrotte et al., “Beam dynamics studies for the fault tolerance assessment of the DPS-XADS linac
design”, 4th OECD NEA International Workshop on Utilization and Reliability of HPPA, May 2004,
Daejon, S. Korea.
[10] J. Galambos, “Operational Experience of a Superconducting Cavity Fault Recovery System at the
Spallation Neutron Source”, these proceedings.
[11] P. Pierini, “ADS Reliability Activities in Europe”, 4th OECD NEA International Workshop on
Utilization and Reliability of HPPA, May 2004, Daejon, S. Korea.
[12] P. Pierini, “Reliability Studies for a Superconducting Driver for an ADS Linac”, 5th OECD NEA
International Workshop on Utilization and Reliability of HPPA, May 2007, Mol, Belgium.
343