Modelling and Monitoring Geomagnetically Induced Currents in Ireland

Abstract

This thesis is the first detailed study of the effects of geomagnetic storms and geomagnetically induced currents (GICs) on the Irish power network.

In order to better monitor geomagnetic storms in Ireland, a network of geomagnetic observatories (MagIE) was set up to complement Met \'Eireann's Valentia observatory. Two permanent magnetometer installations with realtime communications were established in Birr and Armagh to improve realtime geomagnetic coverage. Each site calculates local K-indices in near-realtime. These K-indices, along with the MagIE data can be found at www.RosseObservatory.ie. Geolectric fields were also measured in Sligo and Leitrim at temporary installations. A test magnetometer installation identified Malin Head as a suitable site for a future addition to the MagIE network.

In order to estimate how geomagnetic disturbances drive GICs in the Irish power network, a detailed high-voltage power network model was constructed. This model takes into account the types of transformers and their winding resistances. In addition, some of the substations have associated grounding resistance information, and 15 transformers have ground switches. This detailed model was subjected to a battery of tests to identify substations which are particularly susceptible to GICs. The 400 kV Moneypoint substation in the West of Ireland was found to be by far the most susceptible substation to GICs.

GICs were simulated in the Irish power network model for different recent storms using the spherical elementary current systems (SECS) method of geomagnetic field interpolation, and both the MT and thin-sheet methods of calculating surface electric fields. The detailed power network was imposed on the electric fields and GICs were calculated at each substation. The resulting GICs were compared to measured GICs at a single transformer. It was found that the MT method of calculating electric fields coupled with a uniform 400 $\Omega$m resistivity model gave the most accurate calculated GICs.

Using this model, the March 1989, November 1991 and October 2003 storms were simulated. A maximum GIC value of 168 A was calculated for the Moneypoint substation for the March 1989 geomagnetic storm.

Finally, an estimation of a 1-in-100 year GIC event was calculated using two methods. The first used a previously calculated estimate for a 1-in-100 year geomagnetic storm. This value was used to scale an historic storm, and a maximum GIC value of 255 A was calculated for Moneypoint. The second estimate was made by calculating 25 years of GICs from 1991 to 2015, and fitting both a power law and lognormal distribution. These fits were extrapolated to estimate a large event. The power law gave an estimate of 258 A. The same value for the lognormal distribution was a more conservative 178 A.