1
Implications of Fault Current Limitation for Electrical Distribution Networks Steven M. Blair, Andrew J. Roscoe, Campbell D. Booth, Graeme M. Burt Institute for Energy and Environment, University of Strathclyde, Glasgow, UK [email protected] Anita Teo, Chris G. Bright Strategic Research Centre Rolls-Royce Plc, Derby, UK [email protected] The 10 th International Conference on Developments in Power System Protection, Manchester, 29 th March – 1 st April 2010 The future need for fault current limitation Case study network Scale to number of UK substations Establish a generation scenario that satisfies an 80% drop in CO 2 emissions by 2050: Determine statistical fault level contribution for future generation mix (based on typical interfaces and their fault current contributions) Account for “independent nodes” Combine normal distributions: provides probability (or percentage) of substations with fault level violations Multiply by the number of substations in the UK, for each voltage level Technical issues with fault current limitation An approach has been developed for estimating the number of UK substations that may require fault current limitation. It is based on the statistical analysis of existing fault level headroom, combined with the potential future presence of distributed generation (DG) and energy storage. % required Total energy required Peak capacity Wind 60 % 22.7 GWa 64.9 GW Tidal 20 % 7.8 GWa 18.6 GW Wave 10 % 4.1 GWa 11.7 GW Solar 5 % 2.7 GWa 22.2 GW Despatchable renewables 5 % 1.9 GWa 3.1 GW Storage n/a 5.1 GWa 20.5 GW Mean number of substations requiring fault current limitation: 11 kV 33 kV All three DNOs 282 42 DNOs A and B 320 75 Statistical analysis of present-day fault level “headroom” (from public DNO documents) During a fault, the superconductor in a resistive superconducting fault current limiter (SFCL) will rapidly transit from the superconducting state to a resistive state. However, up to several minutes may be required for recovery to the superconducting state, which leads to several issues. Remote faults, such as F 1 , may cause undesired operation of an SFCL; the DG must “ride-through” until the appropriate protection operates. Fault occurs Protection operates CB fully open Reclosure CB fully closed Re-sync and reclosure CB fully closed Dead time Dead time and SFCL recovery Time Typical reclosure Reclosure with SFCL recovery Non-fault transients, such as transformer inrush, may cause spurious operation of SFCLs. Hence, SFCLs located at the grid infeed must be rated to operate only for “near” faults: Comparison of reclosure times for a transient fault Potential solutions include: • The use of SFCLs that can recover immediately after a fault, or under load • Network automation could manipulate normally- open/closed points to reduce the fault level • Islanding may help to supply loads during SFCL recovery • Islanding can help keep transformers energised, minimising the risk of SFCL operation due to inrush Conclusions: • Suggests a worldwide market for fault current limitation, for countries facing similar future DG connection • Does not include the potential “intangible” benefits of SFCLs: increased interconnection; increased security of supply; increased power quality; and reduced losses -3500 -2500 -1500 -500 500 1500 2500 3500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 Current (A) -20000 -15000 -10000 -5000 0 5000 10000 15000 20000 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 Current (A) Inrush vs. 11 kV bus fault Inrush vs. 33 kV bus fault Transformer inrush Three-phase to earth fault Auto-reclosure schemes (e.g., after a transient or semi- permanent fault at F 2 ) may be complicated by the presence of SFCLs in the network: the dead time may be unduly increased by an SFCL’s recovery time. G 33 kV 275 kV N/O bypass Local loads SFCL DG DG Grid F 2 F 1 11 kV 11 kV Overhead interconnector DG Current at a grid infeed to the 33 kV bus
