1 ITR/P1-03

Status of the Negative Ion based Diagnostic Neutral Beam for ITER

B. Schunke1, I. Ahmed

2, M. Bandyopadhyay

2, U. Baruah

2, D. Boilson

1, D. Bora

1,

A. Chakraborty2, J. Chareyre

1, H. Decamps

1, F. Geli

1, J. Graceffa

1, R. Hemsworth

1,

A. Krylov3, D. Lathi

1, G. R. Nair

2, A. Panasenkov

3, C. Rotti

2, S. Shah

2, M. J. Singh

1,

N. P. Singh2, M. Urbani

1

1ITER Organization, Route de Vinon sur Verdon – 13115 St Paul Lez Durance – France;

2ITER-India, A-29, GIDC Electronic Estate, Sector 25, Gandhinagar – 382025 India;

3Russian Research Centre, Kurchatov Institute, Moscow, Russia

E-mail contact of main author: beatrix.schunke@iter.org

Abstract. In ITER a dedicated 100keV Diagnostic Neutral Beam (DNB) based on negative ion technology will

be injecting 18-20A of hydrogen to provide helium ash measurements via Charge Exchange Recombination

Spectroscopy (CXRS). The DNB will also be used for Beam Emission Measurements (BES), and probably on-

axis Motional Stark Effect (MSE). Reliable DNB performance predictions are essential input parameters for the

design of the diagnostic systems and also assure the correct characterization of all ITER interfaces to allow safe

operation of the ITER plant. The consolidated design of the DNB, including the updated magnetic field

reduction system, is presented, the expected performance parameters assessed and the duty cycle discussed.

1. Introduction

In ITER a dedicated 100keV Diagnostic Neutral Beam (DNB) based on negative ion

technology will be injecting 18-20A of hydrogen to provide the probe beam allowing helium

ash measurements via Charge Exchange Recombination Spectroscopy (CXRS) [1, 2]. The

CXRS diagnostics will also provide measurements of ion temperatures and other essential

plasma parameters and the DNB will also be used for Beam Emission Measurements (BES).

Recently it has been proposed to envisage on-axis Motional Stark Effect (MSE) as an

additional complementary diagnostics. Reliable DNB performance predictions are essential

input parameters for the design of the diagnostic systems. They also assure the correct

characterization of all ITER interfaces to allow safe operation of the ITER plant.

Figure 1 shows a section cut of the DNB installed in the neutral beam cell, which occupies

most of the northern part of the equatorial level of the tokamak building. The DNB is

positioned adjacent to the Heating Neutral Beam injectors (HNBs) sharing equatorial port 4

with HNB1. The DNB injector vessel is roughly 3.5m by 4m by 10m long, and houses the

beam source, the beamline components and the cryo pumps. The necessary RF power, DC

power, hydrogen gas and hydraulic supply arriving via the transmission line are fed through

the DNB High Voltage Bushing which forms the vacuum feedthrough. The DNB can be

isolated from the tokamak via a Fast Shutter (closing time ~1s, vital for limiting the influx of

tritium dust into the injector) or an Absolute Valve [3]. Interventions on the DNB are to be

done by the Remote Handling (RH) system that is common to HNB and DNB. The interface

with the tokamak is shown below.

The baseline performance parameters of the DNB are listed in Table 1. While the baseline

defines a duty cycle of 1:6 there have been recurring requests from the ITER Diagnostic

Section to extend the duty cycle. In response, assessments have been undertaken to identify

the maximum duty cycle technically achievable by the system and the given interfaces. In the

past the interface with the ITER vacuum vessel, or more precisely, the power loading in the

duct and on the blanket opening had been identified as the limiting factor.

2 ITR/P1-03

FIG. 1. Section cut of the DNB in the NB Cell of the Tokamak Building.

TABLE I: DNB SPECIFICATIONS.

DNB Specifications (H-) Operation

Beam Energy (keV/amu) 100

Neutral Beam Power (MW) > 1.6 (3.6 Accelerated)

Accelerated Current (A) 60

Current density (A/m2) 300 Extracted

Stability Energy settling to 1 % after 5 ms

Current density uniformity ± 10%

Divergence 7mrad, target value <7 mrad

Availability 50% of ITER life time

Modulation 5 Hz square wave

Duty cycle 1:6 (3s ON 20s OFF) Baseline

1:2 target value

2. The DNB Beam Source

Negative ion technology was chosen for the DNB first because of the imposed similarity with

the HNBs as using the same technology optimises the required R&D and the plant logistics

and second because positive ion technology was not able to meet the low divergence

requirement which will allow transport of the beam into the ITER plasma over 22m and

through the narrow opening in the ITER vacuum vessel. Furthermore, full power would not

3 ITR/P1-03

.

FIG. 2 The DNB Beam Source.

be available in H0, due to the formation of the H

2 and H

3 fractions. In addition, the efficiency

of conversion for positive ion beams at 100 keV is only ~20%, and would have led to a large

rejection of power in the ion deflection system.

When adopting the RF source as the reference ion source for ITER in 2007, it had been

acknowledged that achieving the performance required for ITER will rely on a dedicated ion

source R&D programme. Among the known issues that have to be addressed are optimised

uniformity of the beam, sufficient level of extracted current, control of the electron current,

avoidance of uncontrolled breakdowns in the accelerator and control of back-streaming ions

and beam optics issues. It has to be noted that the DNB has to be operated at much more

challenging current densities (50% higher than the HNB) and better beam quality, notably

achieve a lower divergence. This is paramount as source uniformity and beam divergence in

particular have an important impact on the signal to noise ratio of the diagnostics.

The DNB beam source (Figure 2)

consists of two parts, the 8 driver

RF negative ion source and the

100kV accelerator. The RF ion

source is almost identical to the

SPIDER source [4] for which the

engineering design was performed

by the RFX NBTF Team in Padua

based on the ion source developed

by IPP Garching [5, 6]. The source

body was then integrated into the

DNB by ITER-India [7], adapting

the interfaces to fit with the DNB

accelerator, the DNB bushing

(horizontal in the DNB) and the

DNB beam source support and

alignment mechanism. The design

of the source back plate and the

cooling lay-out in the Faraday

shield had also been adjusted.

ITER-India has designed the DNB accelerator, crucial for defining the basis of the beam

quality, and substantial work has been undertaken to optimise the design [8], both physics and

mechanical. The basic beam properties such as beam aiming, beamlet divergence, beam

misalignment, power distribution between and within, beam groups and beamlets, defines the

transmission of the beam produced by the DNB to the ITER plasma, and the deposition of the

beam power within the ITER plasma. They define also the power and power density to the

beamline components and the duct between the injector and the ITER plasma.

The DNB extractor / accelerator - based on a 4 by 4 matrix - is a three grid system consisting

of 1280 apertures, grouped in 16 groups with 80 apertures per beam group. Extensive

computational analysis has been performed to estimate the aperture offset steering constant of

the grounded grid (GG) and the extraction grid (EG), the space charge interaction between the

beamlets and the kerb design required to compensate for this interaction. All beamlets in the

DNB must be focused to a single point in the duct, minimizing the beamlet divergence; this

requires point focussing in both vertical and horizontal direction. To suppress the acceleration

of coextracted electrons, permanent magnets have been incorporated in the extraction grid.

Blinker dumps are used as protection from the co-extracted electrons. The beamlet divergence

4 ITR/P1-03

FIG. 3 The beam path through the DNB Beamline

Components.

has been calculated to be 4 mrad. But while the codes used for the accelerator design

reproduce the beamlet divergence reasonably well, the overall divergence of the beam will

also be influenced by mechanical and machining errors and can only be confirmed

experimentally.

3. The DNB Beamline Components

The beamline components (BLCs) consist of the neutraliser, a high density gas cell which

neutralizes the ion beam produced by the beam source, the residual ion dump (RID), which

removes the remaining ions out of the beam and the calorimeter, which is used for measuring

the beam energy and the beam profile during commissioning. When designing the DNB

accelerator, the interception of the beam with the beam line components had also been

studied. The entrance and exit apertures for the neutralizer and RID had been optimised for

maximum transmission and minimum power loading. The placement of the BLCs has been

made after a detailed study of the pressure profile that ensures maximum transmission.

As the DNB for ITER has to deliver beam pulses of up to 3600 seconds, the beamline

components have to be actively cooled so that the rate of heat removal balances the rate of

heat deposition into each component. The duty cycle limit of the BLCs had been assessed

during the final design phase prior to the signature of the DNB beamline procurement

arrangement. Detailed thermomechanical and thermohydraulic analyses had been performed

to assess the capability of the BLCs, taking into account the interface constraints such as the

cooling capacity. The main unknown being the beam divergence a parametric study was

performed including the worst case for each beamline component (10mrad for the neutraliser

and the RID, 3mrad for the calorimeter).

Continuous modulation would be

extremely important for the diagnostic

systems as it allows to account directly for

two major sources of error in the

measurement: the effect of background

emission on the measured helium line

emission, which yields the helium ash

concentration, and the effect of the

background emission on the measured H-

alpha beam emission, which is cross

referenced to obtain absolute He ash

measurements. This is of particular

relevance at the center of the plasma,

where beam attenuation is highest and the analysis the most challenging.

A longer duty cycle has also advantages for the beamline thermal cycling as the large thermal

transients associated with the beginning and ending of the three-second period are eliminated.

The modeling by ITER-India demonstrated that the BLCs could operate ∼13 out of 23s

while the blanket module thermal limits, even at best beam transmission (minimum

divergence/maximum power density), would withstand only up to 50% duty cycle with

sufficient safety margin. The re-assessment of the lifetime limits, namely the fatigue life of

the components will be undertaken once the divergence has been confirmed; only then can the

duty cycle limit be confirmed with confidence.

In addition to the DNB inherent limits, two operational limitations to the duty cycle have been

identified as the performance parameters of the DNB have to be compatible with plasma

control and performance. As the DNB will be injecting hydrogen at all times, the duty cycle

5 ITR/P1-03

FIG. 4 The new DNB ACCC configuration.

might be limited if the continuous injection leads to ‘’hydrogen poisoning’’ in the DT phase

thus diluting the fuel mix and impacting on the fusion performance of ITER. This will be

managed by operating instructions, which will set the injection windows via interlocks. The

operating instructions will also manage the shinethrough limit, which requires a minimum

plasma density in the ITER machine to allow safe beam injection without risking any damage

to the inner wall (power limit P<0.5MW/m2).

The BLCs will be equipped with embedded thermocouples and water flow calorimetry to

derive the beam characteristics and allow monitoring of the beam quality during the long

pulses. A simulation by the Kurchatov Institute indicated that the beam vertical aiming and

the beam divergence can be determined with an accuracy of approximately ± 1 mrad.

4. The DNB Magnetic Field Reduction System

In the area of the DNB injector the stray field from ITER ranges in magnitude from 350G to

150G for normal plasma scenarios. In order to avoid deflection of the charged particles and

optimise the beam divergence this field has to be reduced particularly between the exit of the

accelerator and the neutralizer (Table 2). This reduction is achieved with the Magnetic Field

Reduction System consisting of the Active Correction and Compensation Coils (ACCCs) and

the Passive Magnetic Shield (PMS). The function is twofold – reduce the residual field inside

the injector due to the stray magnetic field from the tokamak to meet the stringent residual

field requirements for optimum beam performance - and compensate the effect the mass of

the injector itself (around 1200 tons of iron and steel) has on the magnetic field of the

tokamak.

TABLE II: DNB RESIDUAL FIELD REQUIREMENT.

Residual Magnetic Field Requirement in the DNB

Gap between Beam Source and Neutralizer |Bz| < 0.5 10-4

T and |∫ Bz dx| | < 0.5 10-4

Tm

Neutraliser |Bz| < 0.2 10-4

T and |∫ Bz dx| | < 0.6 10-4

Tm

Beam source |B| < 1.0 10-3

T

Recently the sets of nested coils

above and below the injector have

been replaced by 2 sets of 3 coils of

equal size, easier to manufacture and

install (Figure 3). The new coil

configuration reduces the weight of

the individual coils and allows

replacement of a faulty bottom coil

by remote handling. As nevertheless

such an operation is complex and

time consuming, double pancakes are foreseen to reduce the need for this rescue scenario.

In addition to serving as a passive magnetic shield, the PMS also assists in attenuating the

neutrons to reduce the level of activation in the NB cell. The single thick soft iron layer has

been replaced by three 50mm layers separated by 2 air gaps of 50mm allowing to minimise

the ACCC current requirement.

6 ITR/P1-03

FIG. 5. DNB PS Modulation Capability.

FIG. 4 The cross-over duct.

5. The interface with the ITER Vacuum Vessel and the duct

Originally injecting radially, the DNB injection angle was modified to 6 degrees in 2007 to

avoid ripple trapped particles and consequently significantly reduce the risk of high heat loads

on the wall below the injection point. The increased CXS background would have also

perturbed the diagnostic measurements. The DNB shares the equatorial port 4 with HNB1,

and the beam is delivered to the ITER plasma through a narrow blanket aperture of 40cm (H)

× 45cm (W), at a distance of ∼22m from the centre of grounded grid.

Based on recent JET results [9], the

estimates for the atomic flow from the

tokamak plasma into the NB duct have been

revised for some of ITER’s high fusion

performance plasma scenarios. The

maximum gas flow is now evaluated to be

4 1020

atoms/m2s (assuming 50% T and

50% D) - a factor 20 higher than previously

envisaged. The duct design has to take into

account these high edge densities and the

HNB duct liner design has been updated

accordingly to cope with the increased

power loads. Detailed calculations of gas density and re-ionised power load profiles along the

DNB line are being carried out with the new atomic flow from the tokamak. Any beam

interception with the surface will lead to the generation of particles from the dumped beam

and desorption from the duct walls, at a position where the injector pumps would not be able

to remove the gas which would freely flow into the tokamak. This would maximise the beam

loss by re-ionisation, increasing the beam losses from direct interception. Because of the

lower beam energy, the DNB will be more significantly affected by beam blocking so that the

power delivered to the plasmas for these high edge density scenarios could be reduced (by up

to 20%) impacting on the S/N of the diagnostics [10]. This places a higher degree of emphasis

on the need for a low divergence (and a higher current density) and is a matter of specific

importance in the planned R&D in the INTF.

6. The DNB Power Supplies

The main components of the DNB Power

Supply System are housed in Buildings 37

and 34. The required controllable voltages

and currents (-96kV dc @ 75A for the

acceleration grid, -12kV dc @ 140A for the

extraction grid) and the 1MHz, 800kW for

the 8 RF antennas, are fed via a ∼120m long

transmission line to the injector in the NB

cell, which enters the tokamak building at

the L3 level. While the plasma production in

the RF ion source is steady-state, the ON-OFF waveform is obtained by modulating the

accelerator via the Acceleration Grip Power Supply (AGPS). The DNB power supply has

been conceived to perform 3600s continuous 5Hz operation (Figure 5); it is not limiting the

duty cycle. While generally the DNB PS is expected to operate with a pre-programmed

waveform, event triggers are also foreseen. The system will be designed to allow stable

operation in the harsh electromagnetic environment of ITER, e.g. cope with fast power

7 ITR/P1-03

switching transients generated within the system and the stray magnetic field from the

tokamak. In collaboration with the diagnosticians the requirements for the control system start

and rise time listed in Table 3 have been agreed to allow efficient diagnostic tuning. The

possibility of 10Hz operation had been discussed, but abandoned in favour of longer flat-top

phases.

TABLE 3: DNB MODULATION SPECIFICATION: TARGET VALUES

Jitter ±1% of start & stop time of each

modulation, i.e. length of modulation ON

time is accurate to within ±1ms.

Beam settling time <10 ms

Waveform square wave – modulation from fully ON

to fully OFF

On/Off Duty Factor for Modulation 50% ON / 50% OFF

Beam turn off time <1 ms

Energy accuracy ±1%

Energy stability ±1%

Combination of accuracy and stability ±1.4%

Current stability ±1.5% after settling time

7. Availability of the DNB – Readiness for operation

As the required performance parameters for the ITER neutral beams are far off today’s

values, a very challenging R&D route is to be undertaken via the Neutral Beam Test Facility

(NBTF) and preparatory ion source experiments in Garching. While the dedicated ion source

testbed SPIDER (part of the NBTF) is committed to developing the ion source for ITER, it

has to be acknowledged that experimental time will be limited and the programme likely to be

centered on the HNB requirements. As the DNB performance requirements are even more

challenging, it is likely that this task cannot be completed in the SPIDER facility. ITER India

has undertaken a voluntary task to demonstrate the DNB performance in India in a dedicated

DNB testbed which will allow to bridge the gap between the SPIDER performance and the

required DNB performance; indeed the demonstration of the DNB performance is part of the

acceptance test for ITER. The INTF provides a unique feature – a far-field measurement of

the beam profile which will allow to measure the divergence and thus allowing reliable

predictions of the DNB power actually delivered to the plasma. This parameter is essential for

obtaining dependable signal to noise estimates needed for the design of the diagnostic

systems. As SPIDER and DNB are now pursuing different accelerator designs, the testing

phase in the INTF has become even more important. Procurement of the DNB beam source,

part of the DNB power supplies and the proto-type DNB BLCs, which form the primary

hardware for the INTF, starts in 2012. The procurement process for the 9m long INTF vessel

has been initiated and the INTF building premises are already available.

8 ITR/P1-03

Figure 6: Shine through limit in H.

Installation of the DNB injector in ITER will begin as soon as the tokamak building is

available to allow sufficient time for commissioning and debugging. Installation and access

logistics require that the HNBs and the DNB are installed in parallel, with the same

infrastructure and common tools being used

sequentially. The NB commissioning is largely

independent of the tokamak activities – the

DNB has a calorimeter to perform the off-line

commissioning even though it is expected that

tight time constraints will make the preparatory

work in the INTF essential. Only duct

commissioning necessitates suitable plasmas,

that meet the minimum density required to

avoid shine through. The low beam energy

makes operation of the DNB possible even at

the low plasma densities achievable with 2MA

current. The DNB should therefore be

available for all ITER operating scenarios from

the hydrogen phase onwards, shine through

permitting (Figure 6).

8. Summary and outlook

With both procurement arrangements covering the DNB signed and the first contracts for the

lead items in the tender phase, the DNB is well on its way. ITER-India will be operating

ITER equipment in the Indian Test Facility (INTF) for extended testing of the DNB beam

source. The demonstration of the DNB performance in the INTF will be an integral part of the

acceptance testing, mitigating the risk that the ambitious DNB performance parameters cannot

be met. Operation in the INTF will also reduce conditioning and commissioning time in

ITER. The experimental measurements of the beam divergence allow to confirm the duty

cycle limit with confidence. The DNB injector has been designed in such a way that the duty

cycle will not be limited by inherent parameters; limitations will be imposed by the interfaces

notably with the vacuum vessel and the operational regime. The DNB will be available as a

diagnostic tool in the hydrogen/helium phase shine through limits permitting.

References:

[1] CHAKRABORTY, A., et. al., IEEE Transactions on Plasma Science, 38 (2010) 248

[2] SCHUNKE, B., et. al., Proc. NIBS, 2009, Vol. 1097 (2009) 480

[3] URBANI, M., et. al., Proc. 27th SOFT, Liege, Belgium, 24-28 Sept. 2012, to be published

[4] MARCUZZI, D., et al,. Fusion Engineering and Design 82 (2007) 798

[5] SPETH, E., et. al, Nucl. Fusion 46, (2006) 220

[6] KRAUS, W., et. al., Rev. Sci. Inst., 79 02C108 (2008)

[7] SINGH, M. J., et. al., Fusion Engineering and Design 86 (2011) 732

[8] SINGH M. J., et. al., Rev. of Sci. Instruments 81 (2010) 13305

[9] LOARTE, A., et al., Proc. 22nd IAEA Fusion Energy Conference, Geneva, Switzerland, 2008,

paper IT/P6-13

[10] SINGH, M. J., Proc. NIBS, Jyväskylä, Finland, 3-7 September 2012, to be published

Disclaimer

The views and opinions expressed herein do not necessarily reflect those of the ITER

Organization.