JET C(98)69 FG Rimini and e JET Team DT Fusion ower roduction in ELM-free H-modes in JET
This document is intended for publication in e open literature. It is made available on e understanding at it may not be furer circulated and extracts may not be published prior to publication of e original, wiout e consent of e ublications Officer, JET Joint Undertaking, Abingdon, Oxon, OX14 3EA, UK. Enquiries about Copyright and reproduction should be addressed to e ublications Officer, JET Joint Undertaking, Abingdon, Oxon, OX14 3EA.
JET C(98)69 DT Fusion ower roduction in ELM-free H-modes in JET The JET Team 1 presented by F G Rimini. JET Joint Undertaking, Abingdon, Oxfordshire, OX14 3EA, 1 See appendix to The JET Team presented by M L Watkins. 83
ABSTRACT Experiments in e ELM-free Hot-Ion H-mode regime have been carried out in Deuterium- Tritium plasmas in JET. Initial experiments undertaken at constant NB power, around 11 MW, demonstrated at core fuelling was dominated by wall/target recycling raer an NB fuelling and made it possible to arrange an optimum core DT mix by adjusting e DT mix in wall/target, gas and NB. High power experiments at 4.2MA/3.6T have successfully and reliably delivered fusion power ( DT ) up to 16.1 MW and plasma stored energy (W DIA ) of 17 MJ. The results are in good agreement wi extrapolations, carried out wi TRANS and JETTO codes, from similar deuterium discharges. Transiently, values of Q around.9 ±.1 were achieved, consistent wi values of n DT () τ E T i () ~ 9 x 1 2 m -3 s kev. The ratio of fusion power to input power Q IN is in excess of.6. There are indications of an isotope effect on e edge pressure pedestal, however no net dependence of global confinement on isotopic plasma composition has been found. 1. INTRODUCTION The first pilot experiments using 1% tritium in deuterium in e ELM-free H-mode at JET produced 1.7MW of fusion power as reported at e 1992 IAEA [1].Thereafter e JET programme turned to e study of two divertor designs of increasing closure. The impact of ese changes on e ELM-free regime in deuterium were described at e IAEA meetings in Seville 1994 [2] and Montreal 1996 [3]. Since e Montreal meeting, e properties of ELM-free Hot- Ion H-modes wi combined Neutral Beam (NB) and Ion Cyclotron Resonance Frequency (ICRF) heating at e Hydrogen minority resonance have been extensively explored. At e same time, e understanding of e MHD stability of e regime progressed wi e identification of e so-called Outer Mode as an ideal external kink, driven by e edge current gradient. A meod for mitigation and avoidance of e Outer Mode by decreasing e plasma current during e ELM-free phase was successfully developed. As a result e performance, robustness and reliability of e regime have been improved, extending e DD neutron yield to 5.2 1 16 s -1 at 3.8 MA/3.4 T wi combined NB+ICRH, ereby making e regime an ideal candidate for experiments in DT. 2. DT MIXTURE CONTROL AND ARTICLE TRANSORT EXERIMENTS In ELM-free H-modes in deuterium e plasma density rise is of e same order of e NB fuelling, but it is not possible to establish e relative contribution to e plasma fuelling from e beams and e wall recycling sources. In e DT mixture control experiments e beam mix was varied from 1 % deuterium to 1 % tritium at constant power, in e range of 11 MW, wi deuterium recycling and gas fuelling. The TRANS code was used to predict, on e basis of a pure deuterium discharge, e DT fusion power as function of e plasma and NB DT mix: For each NB mix, e fusion power 85
8 DT was en computed for various values of Walls pure D e contribution f W of recycling, assumed to be Wall source φ D W = T fw (φ + φ D ) beam beam pure deuterium, relative to at of NBI fuelling. The set of curves us obtained can be 42656 6 f w = compared wi e measured DT (Fig. 1). Note.4 42657.8 at e 1 % T beams case is e most sensitive to f W. In Fig. 1 e best match to e ex- 2. 4 1.2 1.6 42647 perimental data is found for f W = 2, indicating 2 at wall recycling plays a considerable part in Curves: predictions for 11MW NBI e plasma composition in e ELM-free regime, even in low recycling conditions. The Experimental data.2.4.6.8 1. T φ insights from ese experiments made it possible to arrange an optimum core DT mix by pre- Fig. 1 : predicted and achieved fusion power DT for beam / (φ T beam + φ D beam ) ELM-free Hot-Ion H-modes wi ~ 11 MW NBI only, as paring e recycling surfaces to achieve e function of beam DT mix and recycling parameter f W. desired isotopic mix. The spatial variation between D and T sources in e DT mixture experiments allows a more detailed analysis of particle transport by e TRANS code under e assumption of a radially constant diffusion coefficient D o. Additionally, a common local convection velocity is imposed on bo species, sufficient to satisfy e al hydrogenic particle balance. Wiin e uncertainties of e TRANS transport model, e study shows at is series of discharges can be described consistently by e same particle mixing model in which e flow of each species is driven mainly by its own gradients; e value of e core particle diffusion coefficient is of e order of e effective heat conductivity χ eff The assumption of D o = const. is, however, not unique. A functional form of D o which increases towards e edge could also be adopted, as was done to describe tritium transport in transient gas puff experiments in JET ELMy H-modes [4]. 3. OVERVIEW OF HIGH OWER EXERIMENTS The DT experiments in e Hot-Ion ELM-free were carried out first in e standard 3.8 MA / 3.4 T scenario, bo wi NBI alone and wi NBI + ICRH. The NB only case reached 12.3 MW of fusion power wi 21.6 MW of input power, while e combined heating case delivered 12.9 MW of fusion power for 22.6 MW of al input power, close to at expected on e basis of DD performance if e effect of alpha heating is included. In bo cases e Tritium concentration was of e order of 5 % in e core. A second set of experiments was performed at higher plasma current and toroidal field, 4.2 MA / 3.6 T, to exploit e full NB power capability and e expected improvement in confinement and ELM stability wi plasma current. A significant gain in performance was us achieved, resulting in a new record in fusion power, 16.1 MW, and stored energy, 17 MJ, for an input power of 25.7 MW. The ermonuclear fusion power reaches 6 % of e al DT fusion power. No alpha particle driven Toroidal Alfvèn Eigenmodes (TAE) were observed, which is consistent wi eoretical analysis [5]. 86 fus (MW) JG98.62/1c
The achieved level of DT performance Thermal confinement time depends in part on e existence of isotope effects on confinement and MHD stability. In e DT experiments performed to study alpha particle 1.5 1. heating [6] an increase of e density, ion and electron temperatures close to e edge.5 transport barrier wi increasing isotopic mass was measured [7]. In is set of data, however,. Thermal (el.+ion) pressure R = 3.75 m e global confinement time does not vary wi plasma composition (Fig. 2). In e high power 6 ulse No. 42656 experiments, wiin 1 % e loss power is 5 T -> D found to be e same function of stored energy ulse No. 4287 in DD and DT. On e oer hand, local confinement analysis suggests at in e core e 4 3 effective ermal conductivity follows gyro- 2 2. 2.5 Bohm scaling and hence degrades wi increasing mass 3. mass. The fusion power is close to what is predicted by e transport code TRANS, including Fig. 2 : al ermal confinement t E and ermal pressure, at R=3.75 m, as function of edge isotopic compo- alpha particle heating but excluding sition. ELM-free Hot-Ion H-modes wi ~ 1-11 MW NBI only. isotope effects. The experimental maximum net isotope effect maybe of e order of 1 MJ. In summary, isotope effects on local confinement have been identified, bo in e core and in e edge transport barrier, but e net effect on global confinement is small wiin e uncertainties of e analysis. 4. DT FUSION ERFORMANCE ARAMETERS The ermal and al fusion gain can be defined as: (s) r (ka) JG98.62/3c Q DT DT = and Q = rot fast + + Loss αabs Loss Loss Loss αabs where DT and DT are e ermal and al fusion power. The ermal loss power is abs + α, abs dw / dt ; rotation and fast particle components of e loss power are defined analogously. Alpha particle slowing down is also taken into account. The value of e instantaneous power gain is Q IN = DT / IN. It is found at Q increases wi plasma current and is similar for NB+ICRH and NB only discharges. In e record fusion pulse at 4.2 MA/ 3.6 T (Fig. 3) Q and Q initially increase rapidly, en remain fairly constant, us suggesting at fusion power and losses change in proportion. Wi e appearance of MHD activity, clearly seen in e D a trace, bo Q and Q decrease togeer wi e global and ermal confinement times. On e oer hand, e 87
value of Q IN increases monotonically wi time up to e Giant ELM. These observations may indicate at e MHD activity in e ELM-free Hot-Ion regime affects plasma losses more an neutron rate or core conditions. All e DT pulses have been limited by MHD events. However, TOT were it possible to maintain e same confinement quality, W 2 TOT LOSS, and e same reactivity W 2 into steady-state, one would expect e value of Q DT TOT IN to approach e measured value of Q. An overview of e local fusion performance and e input power profiles, as computed by e TRANS code, is given in Fig. 4 for e record fusion pulse just before e MHD activity appears. The data show at, in e core, e fusion power density is of e same order or slightly in excess of e input power density to e ermal component from external sources, ~.6 MW/ m 3. At e same time, e source fusion alpha particle power density is ~.8 MW/m 3, of which ~.6 MW/m 3 is coupled to e electrons; is is comparable to each of e oer electron heating sources, including ion-electron equipartition and change dw e /dt in electron energy density. ulse No: 42976 4.2 MA/3.6T t = 13.25s TRANS Y95 (MW) τ (sec) 3 2 1 1..5 1..5 ulse No: 42976 TRANS run Y935 in loss Q Q heat loss Q in τ E abs τ E dia αabs D α + T α (a.u.) 12. 12.5 13. Time (s) 13.5 14. Fig. 3 : TRANS : time dependent performance and confinement analysis for ulse No. 42976. abs and heat are e power from external sources and e al power, ext. + alpha, absorbed by e ermal plasma. JG98.661/2c (MW/m 3 ) (MW/m 3 ).6.5.4.3.2 Ion input from ext. sources.1 Electron input from ext. sources Electron power input.8.6.4.2 RF Alpha heating 3. 3.1 3.2 NB Ion electron equipartition Fusion power Total input to ermal plasma from ext. sources dwe/dt 3.3 3.4 3.5 3.6 3.7 3.8 Major radius (m) Fig. 4 : TRANS input to ermal plasma, fusion and alpha particle heating profiles for #42976 at t =13.25s JG98.62/7c 5. SUMMARY AND CONCLUSIONS Experiments in e ELM-free Hot-Ion H-mode regime have been carried out in Deuterium- Tritium plasmas in JET. Initial experiments undertaken at constant NB power, around 11 MW, demonstrated at core fuelling was dominated by wall/target recycling raer an NB fuelling; as a consequence an optimum core DT mix could be arranged by adjusting e DT mix in wall/ 88
target, gas and NB. Modelling of particle transport in discharges wi different fractions of T beams into a deuterium plasma has been carried out wi e transport code TRANS. High power experiments at 4.2 MA / 3.6 T have successfully and reliably delivered DT fusion power up to 16.1 MW and W DIA of 17 MJ, and e results are consistent wi extrapolations carried out wi e TRANS and JETTO codes on e basis of deuterium discharges. There are indications of isotope effects local confinement in e core and in edge transport barrier region, but e net effect on global confinement is small wiin e uncertainties of e analysis. Transiently, values of Q around.9 ±.1 were achieved, consistent wi values of n DT () τ E T i () ~ 9 x 1 2 m -3 s kev. The ratio of fusion power to input power Q IN is in excess of.6. In all cases, e high performance phase was transient, limited by external kink modes and ELMs in a similar manner to DD ELM-free H-modes. ower step-down experiments carried out in DD would suggest at Q IN ~Q might be sustained for a confinement time, but even here e performance is ultimately limited by edge stability. It is encouraging to find regimes, such as e Hot-Ion ELM-free H-mode, where e fusion performance can significantly exceed e baseline steady state ELMy H-mode, and such plasmas find ready application in studies of particle and energy transport, alpha particle [7] and ICRF heating and TAE stability. The transient nature of is regime, however, prevents direct application of ese results to a fusion reactor unless means are found of controlling e steep edge gradients, and e resulting edge MHD activity. REFERENCES [1] THE JET TEAM presented by A. Gibson. lasma hysics and Controlled Nuclear Fusion Research (14 Conf. roceedings, Würzburg, 1992) Vol1, IAEA, Vienna (1992) 99. [2] THE JET TEAM presented by. Lomas. lasma hysics and Controlled Nuclear Fusion Research (15 Conf. roceedings, Seville, 1994) Vol1, IAEA, Vienna (1994) 211. [3] THE JET TEAM presented by. Lomas. Fusion Energy (16 Conf. roceedings, Montreal, 1996) Vol1, IAEA, Vienna (1996) 239. [4] THE JET TEAM presented by K.-D. Zastrow, is Conference. [5] S.E. SHARAOV et al., roc. of 25 ES Conf. on Contr. Fusion and lasma hys., 1998. [6] THE JET TEAM presented by.r. Thomas, is Conference. [7] THE JET TEAM presented by V.V. arail, is Conference. 89