implementation of the asynchronous
play

Implementation of the asynchronous landing scheme in KSTAR plasma - PowerPoint PPT Presentation

Implementation of the asynchronous landing scheme in KSTAR plasma control system MinHo Woo 1) , H.Han 1) , S.H. Hahn 1) , J.Kim 1) , Y. S. Bae 1) , J. G. Bak 1) , M. L. Walker 2) , R. D. Johnson 2) , N. W. Eidietis 2) 1) National Fusion Research


  1. Implementation of the asynchronous landing scheme in KSTAR plasma control system MinHo Woo 1) , H.Han 1) , S.H. Hahn 1) , J.Kim 1) , Y. S. Bae 1) , J. G. Bak 1) , M. L. Walker 2) , R. D. Johnson 2) , N. W. Eidietis 2) 1) National Fusion Research Institute, Daejeon 305-333, Korea 2) General Atomics, San Diego CA, U.S.A 1

  2. Outline • Motivation • Conceptual design • Algorithm development • Model validation • Experiment • Conclusion and Future works 2

  3. Motivation Consequences of the disruption  Sudden thermal load to the first wall  Electro-magnetic load on conducting wall  Direct collision of run-away electrons to the wall Cause of the disruption  MHD stability limit in high beta plasma  Hardware failure  Lose of the plasma control Mitigation and avoidance of the disruption  Massive Gas Injection (MGI)  ECCD injection to suppress MHD mode  Asynchronous (soft) landing 2013 KSTAR conference, Lotte Buyeo Resort, February 26-27, 2013 3

  4. Asynchronous landing What is asynchronous landing ?  Switching the poloidal field coil control scheme into the pre-designed ramp down scenario at any time at any plasma current How asynchronous landing works ?  Catch up disruption precursor or hardware fault  Change target plasma current with specified ramp down rate  Feed currents to each PF coil with given ramp rate  Move the plasma inward with fixed plasma shape Advantages of the Asynchronous landing  Reduce the plasma current before hitting the first wall  Reduce the stored energy and eventually wall damage  Same damage pattern for different shots  Enable the tokamak to operate in more dangerous parameter regime without concerning wall damage 4

  5. Conceptual Design • Severe fault  Immediate termination  Ip min fault  Total MVA fault  PS fault • Median fault  Asynchronous landing  PFC fault  rtEFIT fault  External fault … • Intended landing  Locked mode coil  IVC coil  Zp estimator 5

  6. Algorithm Development • Different types of asynchronous landing 1. Intended landing  Monitor precursor signal of the MHD  IVC or Zp to determine if plasma has lost control 2. By hardware faults  External faults by ECH, NBI etc…  PFC armor fault  Ip control error faults • Flow of the sequence  Different sequence has different phase for different category  When asynchronous landing triggered sequence changes and so is phase 6

  7. Algorithm Development • Block Diagram for Asynchronous landing 7

  8. Model Validation • Plasma control simulator - simserver Closed loop simulator with plasma current and shape control with model based on real experiments  Plasma shape Upper single null to circular  Plasma current Ramp down from actual Ip  Plasma position Move inward about 10cm Red line- before landing Black line – after landing Landing started at 5s 8

  9. Experiment shot 8771 Asynchronous landing at 21.375s due to the NBI-1 failure at 21.125s  Change of target Ip and actual Ip  Change of the feed forward current  Drop of the electron temperature  Sudden increase in loop voltage  Elongation recovers to 1.85 2013 KSTAR conference, Lotte Buyeo Resort, February 26-27, 2013 9

  10. Experiment • Robustness of the asynchronous landing Landing by IVC  Strong IVC oscillation before landing  Strong IVC at 5s  PF5 feed forward change from 5s 10

  11. Experiment Landing by PFC fault  High temperature of the Plasma facing component due to the previous experiment  Fault received from external fault monitoring system  Similar behavior to the IVC fault case 2013 KSTAR conference, Lotte Buyeo Resort, February 26-27, 2013 11

  12. Conclusion and Future Works Conclusion • Asynchronous landing scheme is developed • The model is validated using plasma control simulator • It is applied real experiment in 2013 KSTAR campaign • Asynchronous landing works well in various situations Limitations and Future works • Changing feed forward PF current technique is very limited • Problematic when one of the PF coils are saturated • Sudden change in PF request needs to be improved • On-line feed forward scheme is under development • Extend its applicability to more general dangerous situations in KSTAR 12

Recommend


More recommend