SETU solar event to utility

Forecasting geomagnetically induced current in the North East Indian transmission grid, and deciding what to do about it. Every number on this page is produced by the code in this repository, running on real solar wind data from NASA OMNI and real ground magnetometer data from INTERMAGNET.

The standing record

Loading the forecast record.

Storm replay

forecast range, 10th to 90th percentile forecast median what the ground actually did alert level 0.3 nT/s

The forecast at each point uses only solar wind that had already arrived at the first Lagrange point by that time. The observed line is drawn for comparison and was never available to the model.

Regional exposure at the selected time

low elevated high circle size is current per phase per transformer

Most exposed transformers

Warning time

A disturbed period is a run of minutes above the level, separated from the next run by at least two hours. Warning is counted from the last alarm that was already standing when the period began, plus the forecast horizon itself. Periods with no standing alarm are counted as missed and are shown here rather than dropped.

Recommended action

Where to put blocking devices

Sites are chosen against a hypothetical extreme event, not an observed storm. The right hand column is what a ranking that scored each site on its own would have said, which is a different answer.

How it works, in four steps

Four panels showing the solar wind, the ground magnetic disturbance, its rate of change, and the induced electric field through the May 2024 storm
Every panel is drawn from measured data. The solar wind is on the clock of the spacecraft that recorded it, so the gap between the first panel and the second is the warning time the system has to work with.

Why this region

Surface impedance and apparent resistivity of four ground models
The Shillong Plateau returns about seven times the electric field of the Bengal basin for the same magnetic disturbance.

Why the direction matters

Polar plot of network exposure against field direction, next to a map of the regional grid
The network is far more exposed east to west, and the storm time currents that reach Indian latitudes produce exactly that direction.

Scope and limitations

In scope. The 400 kV and 220 kV transmission levels of the North East Region, quasi direct current effects of a geomagnetic storm, the chain from solar wind at the first Lagrange point through to transformer consequence, and operator decision support.

Out of scope. Distribution networks below 220 kV, three dimensional magnetotelluric inversion, induction in pipelines and railways, direct control actuation, and security of the data path.

Known limitations. There is no public record of measured transformer neutral current in India, so the consequence model rests on physics and on published international measurements rather than on Indian ground truth. The network model is representative and not exact, because real network parameters are restricted. Shillong and Tirunelveli observatory data are held by the Indian Institute of Geomagnetism and a data request is pending, so the model is trained on Alibag and Hyderabad.