Reducing Rapid Wind Farm Power Fluctuations Using the Modular Multilevel Converter

Transcription

1 Reducing Rapid Wind Farm Power Fluctuations Using the Modular Multilevel Converter Abel A. Taffese, Atsede G. Endegnanew, Santiago Sanchez, and Elisabetta Tedeschi Department of Electric Power Engineering, NTNU Sintef Energy Research January, 2019

2 Table of Contents Introduction Outcomes Modular multilevel converter Method Results Simulation Laboratory Conclusions

3 Introduction Outcomes Outcomes of the paper Objective To Develop a tool to smooth the rapid power fluctuations of wind farms with MMC. Santiago S. NTNU January, / 10

4 Introduction Outcomes Problem description Offshore wind farms integration Santiago S. NTNU January, / 10

5 Introduction Outcomes Problem description There are three power fluctuations types that are linked with the wind variability: 1. Long-term seasonal variations. 2. Short-term time scale of minutes to few hours. 3. Rapid changes fast variations (wind gusts, tower shadow,...). Santiago S. NTNU January, / 10

6 Introduction Outcomes Problem description P. Sorensen et. al., Power fluctuations from large wind farms, IEEE Trans. on Power Systems, vol 22, no. 3, 2007 Santiago S. NTNU January, / 10

7 Introduction Outcomes Problem description Since short-term and rapid changes are difficult to predict, energy storage solutions are being proposed to smooth the variations out. Santiago S. NTNU January, / 10

8 Introduction Outcomes Problem description Multiterminal for wind farms integration. BP BP Santiago S. NTNU BP January, / 10

9 Table of Contents Introduction Outcomes Modular multilevel converter Method Results Simulation Laboratory Conclusions

10 Modular multilevel converter Capability of an MMC to store energy (W) idc Leg Submodule vdc 2 Cpole vua vub arm vuc Ci idc iua Ra Ra ila va La La iub Ra vb Ra La ilb iuc Ra La Ra ilc vc La La Lf vg vdc 2 Cpole vla vlb vlc vsmi + - Ci + Submodule Dynamic equivalent ; The aim is to control W in order to smooth the fast power fluctuation. - vci Santiago S. NTNU January, / 10

11 Modular multilevel converter Capability of an MMC to store energy (W) source idc P flow Leg Submodule vdc 2 Cpole vua arm vub vuc Ci Level sink idc iua Ra Ra ila va La La iub Ra iuc Ra La La vb Ra La ilb vc Ra La ilc Lf vg + vdc Cpole 2 vla vlb vlc vsmi + - Ci vci - Submodule Dynamic equivalent The aim is to control W in order to smooth the fast power fluctuation. Santiago S. NTNU January, / 10

12 Modular multilevel converter Capability of an MMC to store energy (W) source Sink AC power MMC P flow idc Leg Submodule vdc 2 Cpole vua arm vub vuc Ci Level sink idc iua Ra Ra ila va La La iub Ra iuc Ra La La vb Ra La ilb vc Ra La ilc Lf vg + vdc Cpole 2 vla vlb vlc vsmi + - Ci vci - Submodule Dynamic equivalent The aim is to control W in order to smooth the fast power fluctuation. Santiago S. NTNU January, / 10

13 Modular multilevel converter Capability of an MMC to store energy (W) source Source DC power MMC P flow idc Leg Submodule vdc 2 Cpole vua arm vub vuc Ci Level sink idc iua Ra Ra ila va La La iub Ra iuc Ra La La vb Ra La ilb vc Ra La ilc Lf vg + vdc Cpole 2 vla vlb vlc vsmi + - Ci vci - Submodule Dynamic equivalent The aim is to control W in order to smooth the fast power fluctuation. Santiago S. NTNU January, / 10

14 Modular multilevel converter Capability of an MMC to store energy (W) source water level energy storage capability MMC idc P flow Leg Submodule vdc 2 Cpole vua arm vub vuc Ci Level sink idc iua Ra Ra ila va La La iub Ra iuc Ra La La vb Ra La ilb vc Ra La ilc Lf vg + vdc Cpole 2 vla vlb vlc vsmi + - Ci vci - Submodule Dynamic equivalent The aim is to control W in order to smooth the fast power fluctuation. Santiago S. NTNU January, / 10

15 Modular multilevel converter Capability of an MMC to store energy (W) Arm energy dynamics: dw dt = 1 3C (P dc P ac ) (1) W : Arm energy C: equivalent arm capacitor P dc : DC power P ac : AC power The aim is to control W in order to smooth the fast power fluctuation. Santiago S. NTNU January, / 10

16 Table of Contents Introduction Outcomes Modular multilevel converter Method Results Simulation Laboratory Conclusions

17 Method Energy control scheme P ac + V dc,ref V dc ρ P ac,ref + PI Idref Current control V d P ac Santiago S. NTNU January, / 10

18 Method Energy control scheme P ac + V dc,ref V dc ρ P ac,ref + PI I dref Current control V d P ac W ref + W + W control Circulating current I c 3V dc P dc W P ac P dc + 1 3Cs W P ac Santiago S. NTNU January, / 10

19 Method Energy control We use a Non-linear control for W. I c,ref = 1 u c ( d(wref + W ) dt ) + P ac K (W ref + W W ) (2) K : virtual gain. Santiago S. NTNU January, / 10

20 Method Energy control We use a Non-linear control for W. I c,ref = 1 u c ( d(wref + W ) dt ) + P ac K (W ref + W W ) (2) K : virtual gain. V 2 dc T w s T w s+1 w max V 1 dcf 2 g T f s+1 W Remove DC component Low pass filter w min Santiago S. NTNU January, / 10

21 Table of Contents Introduction Outcomes Modular multilevel converter Method Results Simulation Laboratory Conclusions

22 Results Simulation Simulation Test system Power fluctuation Constant P Converter 1 Converter 2 Line 12 T1 Line 14 T2 Line 13 Line 24 Converter 3 Converter 4 Droop mode T3 Line 34 T4 Droop mode W. Leterme, et al., A new test system for dynamics and protection studies in EMTP-type software, in 11th IET International Conference on AC and DC Power Transmission, Santiago S. NTNU January, / 10

23 Results Simulation Simulation Start P 1 = ±10MW Fluctuation of Power 3 and 4 Santiago S. NTNU January, / 10

24 Results Simulation Simulation Start P 1 = ±10MW Fluctuation of DC voltage Santiago S. NTNU January, / 10

25 Results Simulation Simulation When smoothening function is enabled the active powers and dc voltage show improvements Santiago S. NTNU January, / 10

26 Results Simulation Simulation The arm energy larger variation ( W enabled). The method is distributed. Onshore converters (3 and 4) arm energy variation. Santiago S. NTNU January, / 10

27 Results Laboratory National Smart Grid Laboratory (Norway) Rated power: 60 KVA Number of sub-modules: 18 sm AC voltage: 400 V DC voltage: 700 V Santiago S. NTNU January, / 10

28 Current, power (pu) Results Laboratory National Smart Grid Laboratory (Norway) 0.07 ic ref 0.06 pac ff i c Step W ref time (s) Santiago S. NTNU January, / 10

29 Current, power (pu) Results Laboratory National Smart Grid Laboratory (Norway) 0.07 ic ref pac ff i c time (s) Santiago S. NTNU January, / 10

30 Energy (pu) Results Laboratory National Smart Grid Laboratory (Norway) wref wav Step W ref time (s) Santiago S. NTNU January, / 10

31 Energy (pu) Results Laboratory National Smart Grid Laboratory (Norway) wref wav time (s) Santiago S. NTNU January, / 10

32 Table of Contents Introduction Outcomes Modular multilevel converter Method Results Simulation Laboratory Conclusions

33 Conclusions Conclusions We developed an energy controller that helps to exploit the energy storage capability of the MMC. We validated the energy control technique in the laboratory. Fast power fluctuations from wind farms can be compensated applying such controller to the MMCs of the. Santiago S. NTNU January, / 10

34 Conclusions Thank you! Questions Santiago S. NTNU January, / 10

35 Full scheme Santiago S. NTNU January, / 10