Skip to main content

European Risk from Geomagnetically Induced Currents

Final Report Summary - EURISGIC (European Risk from Geomagnetically Induced Currents)

Executive Summary:
The main results of the EURISGIC project are:

1) Statistics of the geoelectric field and GIC in Europe based on geomagnetic data in 1996-2008, on the conductivity map of local 1-D ground models and on a simplified description of European high-voltage power grids. The largest 1-min GIC values are a few hundred amperes and occur in Nordic countries. Values up to about 100 A are possible in the Baltic countries and British Isles, but GIC likely remain below 50 A at more southern regions.

2) Estimations of the worst-case GIC scenarios based on extrapolated statistics of magnetic field variations and on simulations of a few extreme solar wind events. A rough rule of the thumb is that a once in 100 years event would be about twice as large as the October 2003 storm, which caused a blackout in Malmö, Sweden. Simulations indicate that if the July 2012 coronal mass ejection had hit Earth then it would have cause a magnetic storm as large or larger than in March 1989 (Quebec blackout event). An event equal to the Carrington storm in 1859 would be even larger, but it is difficult to give a precise estimate.

3) Prototype GIC forecast server using Solar Shield simulations and an empirical method to provide predictions from real-time solar wind data. The code of the GUMICS simulation was parallelised for a much higher computational power, but it is not yet mature enough for routine forecasting.

4) Recordings of GIC at five power grid substations in North-West Russia, at one site of the Finnish natural gas pipeline and recordings of the geoelectric field at three sites in the U.K. The project period was magnetically quiet especially when compared to the previous activity maximum around 2003.

5) Digitised data of the magnetic field from the 1840’s in the UK and of the geoelectric field at Nagycenk, Hungary, in 1957-1993.

6) Visualisation tools for presenting results of statistics and single major events and for providing demonstrations of GIC in a power grid.

The project website (www.eurisgic.eu) provides access to different services and data.

Project Context and Objectives:
Geomagnetically induced currents (GIC) from solar storms pose a risk to the operation of power transmission grids. The European Risk from Geomagnetically Induced Currents (EURISGIC) project (March 2011 - February 2014) aimed at increasing knowledge of GIC hazard and at developing European capabilities for GIC forecasting. The most famous GIC events were the province-wide blackout in Quebec, Canada, in March 1989 and the blackout in the city of Malmö, in southern Sweden, in October 2003. The integration of interconnected and geographically wide power transmission grids in Europe and elsewhere in the world is obviously increasing the GIC risk.

Calculation of GIC in a given power grid is basically a straightforward problem, which has been considered in several previous national studies, for example, in Finland, Sweden and the United Kingdom. However, no continent-wide modelling of GIC had ever been done before this project.

The EURISGIC project had the following objectives:

(1) Construct a model of the European high-voltage power grids applicable for GIC calculations.

(2) Collect a geomagnetic data set of European observatories in 1996-2008 covering one sunspot cycle.

(3) Compile a catalogue of ground conductivity models for calculating the geoelectric field from the geomagnetic field data.

(4) Derive statistics of GIC in Europe in 1996-2008.

(5) Apply long-term magnetic and electric field data sets to estimate worst-case scenarios of GIC events.

(6) Apply simulations of the magnetosphere-ionosphere system and empirical methods to produce GIC test forecasts for Europe based on real-time in-situ solar wind observations.

GIC is a global phenomenon that does not respects national boundaries. To exploit the best knowledge available in the world, the project also had contributors outside of the European Union, as well as an external advisory group representing power engineering and education.

Project Results:
Previous studies on geomagnetically induced currents (GIC) in Europe have focussed on national power grids especially in the Nordic countries and the UK, but EURISGIC investigated the interconnected grids across the continent. This required a straightforward extension of existing methods to be able to deal with a large geographic area (Viljanen et al., 2012). Additionally, geomagnetic data and ground conductivity models (Adam et al., 2012) were compiled for the whole European region. During the first project year, the consortium reached the full technical capability to perform continent-wide studies.

During the second project year, the updated methods were used to derive GIC statistics across Europe. There are three factors contributing to the occurrence of GIC: topology of the power grid, geomagnetic variations and the ground conductivity. The prototype grid model used in this study indicates that all parts of the European grid can in principle experience large GIC if the electric field has no spatial variations. However, the spatial variability of both the magnetic field and the ground conductivity create non-uniform electric fields.

The geomagnetic field and its time derivative are generally largest in the north close to the auroral region. If the ground conductivity were identical everywhere then the electric field would decrease towards the south. The true conductivity is much more complex, which causes an additional spatial variation. However, North Europe is still the most likely area of large electric fields. Especially, there are regions in the north having a small ground conductivity, which tends to further increase the electric field. Based on geomagnetic data of 1996-2008 and the ground conductivity model of 1-D blocks, GIC levels are highest in the Nordic countries (up to a few 100 amperes) and clearly smaller in the Baltic countries, British Isles and South and Central Europe (less than 100 A). The occurrence of the largest GIC amplitudes concentrated on three days in 1996-2008: 15 July 2000 and especially on 29-30 October 2003. These events give thus a reference to extreme storms outside of the period studied. Key results of GIC statistics are available on the public website on an interactive graphical interface. There is also an access to the daily sum of GIC in different parts of the European grids in 1996-2008. This gives a quick-look tool to identify major events in the past.

The GIC statistics covers one sunspot cycle. This gives a comprehensive understanding of the average geographic, diurnal and annual characteristics. However, much longer time-scales must be considered to be able to estimate the extreme limits (Pulkkinen et al., 2012; Thomson et al., 2011). This is a demanding task, since GIC events with major effects on power grids have been rare with the March 1989 and October 2003 geomagnetic storms as the best-known cases. Additionally, large power grids have existed only for some tens of years, which is a very short time compared to the age of the Sun. So we cannot just use GIC recordings, but we must rely on geomagnetic data from which it is possible to estimate GIC in present power grids.

Another challenging task has been to develop a forecast service of GIC in Europe. Magnetosphere-ionosphere simulations using real-time solar wind observations as input produce the input geomagnetic field to the GIC software. Two different codes were applied: the European GUMICS-4 (Janhunen, 2012) and the US Solar Shield (Pulkkinen et al., 2010). A requirement is that the codes must be fast enough to allow calculation of GIC with about a 30-minutes lead-time. Solar Shield produces continuously test forecasts of GIC for North Europe. The parallelised version of the GUMICS-4 code (GUMICS-5) was tested in a supercomputer and it was capable to provide GIC forecasts of a 30-min lead time. However, some errors in the code of GUMICS-5 were detected and are presently fixed by comparisons with the results by the well-established GUMICS-4. We have also faced the physical limitation that the present simulations cannot produce reasonable results at mid-latitudes.

As an alternative approach to simulations, an empirical method by Wintoft et al. (2005) was extended to relate solar wind parameters directly to the 30-min maximum of the time derivative of the ground magnetic field across Europe. This quantity is highly correlated with the electric field, so it also provides a proxy for the GIC activity when measured as the regional sum of GIC at power grid substations. An advantage of the empirical method is its applicability at all latitudes.

Estimations of the worst-case GIC scenarios were derived based on extrapolated statistics of magnetic field variations and on simulations of a few extreme solar wind events (Beggan et al., 2013; Myllys et al., 2014; Ngwira et al., 2013; Pulkkinen et al., 2012; Thomson et al., 2011). A rough rule of the thumb is that a once in 100 years event would be about twice as large as the October 2003 storm, which caused a blackout in Malmö, Sweden. Simulations indicate that if the July 2012 coronal mass ejection would have hit Earth then it would have caused a magnetic storm as large or larger than in March 1989 (Quebec blackout event). An event equal to the Carrington storm in 1859 would be even larger, but it is difficult to give a precise estimate.

Besides using readily available sources of data, the EURISGIC project has two specific data tasks. One is to perform GIC recordings in North Europe, and another is to digitise old analogue recordings. GIC measurements are running at five sites in North-West Russia and their data are available in near-real-time (http://eurisgic.org/). Additionally, continuation of the long-term recording of GIC in the Finnish natural gas pipeline is supplementing this study (http://space.fmi.fi/gic/). In the UK, recordings of the geoelectric field at two sites have started in 2012-2013, and real time data are now available at
http://www.geomag.bgs.ac.uk/data_service/space_weather/geoelectric.html
and on request for scientific study.

A previously started labour intensive campaign to digitise the full set of UK analogue magnetograms was completed and the high-resolution images are publicly available on-line (http://www.bgs.ac.uk/data/magnetograms/). The next stage to process these images in order to extract digital data that can be used for research was then started. The difficulty of this task has become clear, especially for the stormy periods of most interest. However, progress has been made on digitising and analysing the UK data for the Carrington storm and preliminary results became available.

A uniquely long measurement series of the geoelectric field has been maintained at the Nagycenk observatory in Hungary since 1957 covering about five sunspot cycles (Kis et al., 2007). The original analogue data in 1957-1993 have now been digitised into binary files. These data can automatically be resampled to a uniformly spaced 1-min time series. To convert these values into physical units (mV/km), the necessary information is available in observatory year books and as hand-written notes on the original films. This last step is still very laborious, but will be finished until the end of 2014 as a post project effort.

The EURISGIC consortium was active in dissemination of research results in various forums from international conferences to national high-level authorities. The public website (http://www.eurisgic.eu/) was created in the beginning of the project, and it is continuously updated. It also provides a portal to other EURISGIC service and data websites.

Based on experiences in EURISGIC, we can identify a few items concerning improvements of GIC forecasts:
1) Present global magnetospheric simulations would need more accurate descriptions of the magnetosphere-ionosphere coupling especially to predict auroral substorms, which are the most important cause of large GIC.
2) For forecasts with a 1-2 day lead-time, observations of the phenomena on the solar surface are used to estimate solar wind parameters. Reducing uncertainty in estimating arrival times of coronal mass ejections will still require extensive efforts.
3) Modelling of the geoelectric field should be done by full 3-D ground models especially to describe correctly spatial variations of the field at sharp conductivity boundaries. For operative forecasts, supercomputers would then be necessary.
4) Access to real-time ground magnetometer data in Russia could provide some short-term advance warning of increased magnetic activity in Europe. This can work when an active ionospheric current system observed above Russia remains active until the European region experiences it due to Earth’s rotation.

Potential Impact:
POTENTIAL IMPACT

The EURISGIC project has provided the first comprehensive estimation of the GIC occurrence in the high voltage power transmission network across Europe. This helps power grid operators to identify the regions where large GIC values are likely to occur.

The project has demonstrated the capability to provide GIC forecasts with a 30 min lead time based on in situ solar wind observations. For monitoring the present GIC levels, a corresponding real-time calculation can be based on the measured geomagnetic field at ground-based observatories.

As known from historic evidence, extreme GIC events are the most relevant ones having the potential to cause power grid blackouts, which has happened about once in ten years in the past. Although occurring rarely, such events can cause wide-spread impacts across national boundaries in the increasingly more interconnected power grids. In EURISGIC, special attention was paid to estimate the magnitude of a once in 100 years event. Such a time-scale is important when designing new grids or new connections between previously separated grids. Although large GIC values typically occur only in North Europe, comparable currents could be reached in Central and South Europe during a extremely strong magnetic storm. Such an event has never occurred during the presence of modern power grids, but its probability is non-zero, as indicated by the historic observations of the September 1859 storm.

MAIN DISSEMINATION ACTIVITIES

EURISGIC was a research project, so the main channel to disseminate its results were scientific publications in peer-reviewed journals and presentations at international scientific conferences. The most important yearly conference during the project was the European Space Weather Week in Belgium, and it was contributed by all consortium members. This conference collects participants from the scientific and end-user communities. It provided a forum for smaller informal splinter meetings arranged by the EURISGIC consortium. There were also a few topical meetings on GIC arranged by national or European level actors collecting a large variety of audience including scientists, grid operators, civil contingency representatives and insurances.

The project had an external international advisory group whose members participated regularly in project meetings and provided feedback. This also increased awareness and gave more direct contacts to power grid operators in Europe.

The public website of the project (www.eurisgic.eu) presents generally the project and provides a portal to other more detailed EURISGIC service and data websites. GIC data recorded during the project are publicly presented in near-real time.

EXPLOITATION OF RESULTS

The updated software for deriving long-term GIC statistics is straightforward to apply to any national studies anywhere in the world. There was already one example during the project lifetime of such a study made for the Statnett company in Norway. The Space Situational Awareness (SSA) program of the European Space Agency (ESA) provides an immediate further opportunity to continue GIC services (nowcasts, forecasts) at the European level.

Further research is going on to estimate which magnitude an extreme GIC event can reach once in 100 years. These results can be used by power companies to assess the need for possible counteractions or for defining specifications for new transformers.

Measurements of GIC and the geoelectric field can be extended to new locations. Experienced teams can provide tailored recordings to power companies. Although the EURISGIC project focused on studying GIC in power grids, the same methodology is applicable to induced currents in oil and gas pipelines and other long technological conductor systems.

Development of faster space plasma simulations by parallelising codes provides general benefit to other computing applications too.

List of Websites:
Public website:
http://www.eurisgic.eu/

Contact persons at participating institutes:

Finnish Meteorological Institute
Dr Ari Viljanen
ari.viljanen@fmi.fi

British Geological Survey
Dr Alan Thomson
awpt@bgs.ac.uk

NeuroSpace
Dr Magnus Wik
magnus@neurospace.se

Swedish Institute of Space Physics
Dr Peter Wintoft
peter@lund.irf.se

Geodetic and Geophysical Institute
Dr Viktor Wesztergom
wv@ggki.hu

Polar Geophysical Institute
Dr Yaroslav Sakharov
sakharov@pgia.ru

Catholic University of America
Dr Chigomezyo Ngwira
chigomezyo.ngwira@nasa.gov