2.  Can we calculate, measure, predict the GICs?  How will the GICs affect transformers?  How will the power network respond?  What could we have done to reduce damage?  What can we do next? 2 ? ? ? ? ?

3. 3 Transformer real time risk estimate Weather forecast Historical line failure probability Historical trfmr failure probability Real time & alt. network topologies LT network reliability probability analysis Real-time trfmr condition, temp & measured load Line load flows Damage cost model Value-at-risk estimate Planning decisions Long term economic model Customer interruption cost model Capacitor & protection risk model Switching mitigation options Alt. network topologies Line real time risk estimate Operational failure probability analysis Real time GIC measurement Component failure model Credibility test Utility outage cost model Real time GIC calc System Operating Procedure

4. 4 What are the characteristics of a GMD? What will be the driving voltage of the GICs in the electricity network – the E field?

5. 5

6. 6 Projected 1-in-100-year E-field in Quebec: 10-50 V/km (10 sec data) Experience of E-fields: 5-15 V/km Order of magnitude reduction of E-field (1min) at mid-latitudes.

7. 7 Expect past events (Quebec and Halloween storms) to have produced E-fields of 0.3-0.4 V/km at mid-latitudes (1 min data). Consistent with measured GICs in S Africa. Therefore – expect extreme event in Southern Africa to generate 3-4 V/km (1 min), and have effects similar to Quebec storm in North America?

8.  GIC in transformers: distorts magnetic flux, generates harmonics, increases ‘reactive power’, causes overheating, degrades insulation.  Power system failure: stability under non-active power variation.  Transformer condition monitoring: dissolved gas analysis and interpretation. 8

9.  Do nothing - High risk.  Series caps in all lines – High cost; increases transmission capacity, but risk of SSR.  Replace all transformers and reactors with ‘robust’ designs – Unrealistic: time and cost  Change earthing (grounding) of all transformers and reactors – Protection, operations and cost impact.  Implement a temporary network configuration approach – Limited effect in extreme events.

10.  What is the cost of an interruption?  What is the effect on society of a serious dislocation of power supplies?  Who or what provides the ‘cover’ for the damage of an extreme event?  Is a utility responsible for ‘externalities’?  Are costs of mitigation allowable under electricity regulation? 10

11. 11 HE R TSU KMH HBK

12. 12 Calculated from B-field Measured 1-min values NamPower 02 Oct 13

13. 13  Calculation: Four Intermagnet observatories allow B-fields to be calculated with SECS.  Transformer neutral measurements: collaboration with Eskom and NamPower.  Additional MT and magnetometer measurements, but earth resistance mostly unknown. ISSUES  Data base standards, hosting, access.  Adding value by analysis.  Almost no operational prediction.

14.  How does the fringe of the auroral zone change?  Network modelling taking into account the time response of transformers to the change of ‘dc’ shows possibility of resonance … Is there a characteristic frequency of CMEs equivalent to wavelength 1 – 4 min?  Is there a characteristic direction or ‘turbulence’ of induced E-fields?  How accurate is the E-field modelling for distances of 50-300 km? 14

15. Examine GICs in context of risk. Reduce uncertainty within models of GIC: ◦ GMD prediction – onset, magnitude and end of event ◦ equivalent earth surface resistance model, ◦ transformer and system response under unusual conditions. 15