Control and Operation of Multiterminal

Transcription

1 Control and Operation of Multiterminal HVDC (MTDC) and its application for offshore wind energy integration in the North Sea Temesgen Haileselassie, Kjetil Uhlen Department of Electrical Power Engineering, Norwegian Univerisity of Science and Technology ( NTNU) 1

2 Background Why MTDC? Wind power and market integration: Efforts to achieve the 2020 goals (at least 20% renewables by the year 2020) Suitable sites for wind farms are running out on land (or controversial). This has led to looking for new sites offshore. 1,700 MW currently operational offshore wind farms Expected to grow to 19,000 MW by 2015 Going for offshore sites not only offers advantages but also brings about new challenges Grid connection being one of them. 2

3 The grid integration problem in the context of the North Sea 1. Considerations for developing wind farms in the North sea as far as km away from shore HVDC stands out as the more feasible solution for such long subsea power transmission 2. Availability of offshore loads (oil & gas platforms) in the North Sea Trends to electrify oil/gas platforms of Norway from onshore (Troll A, Valhall) by HVDC Multi-terminal HVDC (MTDC) is a promising solution for integration of offshore wind farms, and oil/gas platforms 3

4 4 An early stage scenario of MTDC in the North Sea

5 The European Offshore Supergrid (Proposed by many) VSC-HVDC is the technology to enable such offshore dc super grids. Airtricity Sintef EWEA 5

6 TECHNOLOGY GAPS In general: IGBT technology not widely used for DC transmission Suitability for offshore application must be demonstrated Scaling up from the current capacity level of 400MW XLPE submarine cable systems must be proven for operation at 300kV DC (current projects operate at 150kV) For multi-terminal systems: EHV DC Circuit Breaker Control systems 6

7 Active Power Control in MTDC Each converter should independently control: the active power flow (as given by the power reference signal) the dc bus voltage of the converter to manage the power balance within the DC grid Controls should also: be robust to contingencies contribute to the balance of the ac grids (contribute to the primary frequency droop control) 7

8 Power balance control in AC grids: Traditionally by frequency droop f f f fmax Generation station-1 Pgen1 Generation station-2 Pgen2 Aggregate Ptot 8

9 Power flow control in DC grid : achieved by DC voltage droop UDC UDC UDC UDC,max UDC,min Inverter mode Rectifier mode VSC-HVDC station-1 P1 VSC-HVDC station-2 P2 Aggregate Ptot 9 No need for communication between terminals Many converter terminals contribute to dc voltage regulation DC analogy to distributed frequency droop control in AC systems

10 10 An early stage scenario of MTDC in the North Sea

11 11 Sample view of the DC droop control

12 Frequency support by MTDC Enables sharing of primary reserves between different AC grids connected to same MTDC even though they operate asynchronously. Hence (possibly) lower cost for primary reserves. 12

13 Schematics of the HVDC control AC grid HVDC DC voltage droop controller 13 Frequency droop controller

14 Frequency support - Simulation examples Frequency (pu) Grid-1 Grid Frequency (pu)0.995 Grid-1 Grid-2 Power (MW) Grid-1 Grid-2 Power (MW) Grid-1 Grid-2 14 DC voltage (kv) Time (s) DC voltage(kv) Time (s) Before After

15 Frequency sensitivity analysis of MTDC wrt load changes The objective is to analytically determine interaction of different AC grids connected to the MTDC. Gives a systematic way to determine the dc voltage droop (ρ DC ) and frequency droop (ρ f C) constants. 15

16 Equations for power balances P G P C PLi + PCi = PGi (1) P L f ρi = P i Gi f i P Gi (2) 16 f f U U PCi = PCi ρfci ρdci n j= 1 i i DC DC P = Cj (3) 0 (4) n total number of HVDC stations.

17 After solving (1)-(4) we get the frequency sensitivity expression K1 qp 1 C 2 ρfc 2 qp 1 Ci ρfci qp 1 Cn ρfcn f1pu PL 1 q2 PC 1 ρfc1 K2 q2 PCi ρfci q2 PCn ρfcn f 2pu PL 2 = qp i C1 ρfc1 qp i C 2 ρfc 2 Ki qp i Cn ρfcn fipu PLi q P ρ q P ρ q P ρ K f P Ln n C1 fc1 n C 2 fc 2 n Ci fci n npu Frequency sensitivity matrix 17

18 18 Simulation study of for a four terminal system

19 Terminal No South Norway AC Grid size (GW) Converter size (GW) ρfci (Converter freq. droop) ρdci 12 (6 GW/Hz) Netherlands 6 (3 GW/Hz) England 8 (4 GW/Hz) Offsh. Windfarm (insensitive) (insensitive) UDC=400 kv, ρi = Resulting sensitivity matrix: f P 1pu L f 2 pu PL 2 = f 3 pu P L f P 4 pu L4

20 Simulated shut down of generation (1.64 GW) in the UK ac grid (ΔPG3=1.64 GW) Frequency (pu) Coverter power (MW) Nor. Neder. UK DC voltage (kv) Time (s)

21 Conclusions MTDC has the potential to fully integrate power markets between asynchronous areas. Can be operated in a similar manner as ac grids (with the dc voltage droop control) No need of fast communication between converter terminals, instead DC voltage droop control is used. Primary reserves can be traded between asynchronous areas (with frequency droop on the converter) With primary reserves exchange by MTDC, cost of operating spinning reserves can be reduced. 21

22 Thank you. 22

23 Remarks With the proposed analytical method, the observed and the predicted frequency changes were in good agreement. The method, can further be used for small signal stability study of MTDC connected ac grids. 23

24 24 Comparison of Results Observed ω = P = GW 1pu L1 ω = P = GW 2pu L2 ω = P = GW 3pu L3 Calculated U = DCpu ω = P = 0.203GW 1pu L1 ω = P = 0.110GW 2pu L2 ω = P = 0.314GW 3pu L3 U = DCpu