D1.1 DC description of power grids
Ari Viljanen, Risto Pirjola and Magnus Wik
A model of the European high-voltage power grids is a basic input to the tasks of the EURISGIC project (European Risk from Geomagnetically Induced Currents). Due to the quasi-dc nature of geomagnetically induced currents (GIC), only a DC description is needed. This contains a list of transformer stations (nodes), their locations and resistances, and a list of transmission lines and their resistances. Because GIC is divided uniformly between three phases, a single-phase reduction can be done. This means that resulting GIC are expressed as the sums of the currents in the three phases.
The purpose of the EURISGIC grid model is to provide a prototype which covers whole Europe and has a realistic number of nodes and lines, so that it gives a reasonable understanding of the occurrence of GIC in dierent parts of the continent. The model must also be easy to extend and adjust whenever additional information becomes available.
The European power grid system consists of four parts, which are connected by DC lines. From the GIC viewpoint, these parts can be considered separately. The largest part is the South and Central Europe, and the other parts are the Baltic countries, Nordic countries and British Isles. The number of nodes of the whole model used here is 1487 (1107 in South and Central Europe) and there are 2027 transmission lines (1557 in South and Central Europe). The voltage levels included are 400 kV, 330 kV and 220 kV. This fixed conguration will be the basic input in the later EURISGIC studies.
This document starts with a general overview of the GIC modelling method applied within EURISGIC. Then the construction of the power grid model is described and approximations are explained. Eects on GIC due to varying power grid parameters are discussed by simple examples with comparisons to GIC in the prototype grid model. We have applied the old widely-used Finnish grid model of 1978-79 as a reference, and veried that the basic GIC modelling works reliably with much larger grids too. Applications of the prototype model with real geophysical input data are described in an accompanying document by Viljanen et al. (2012).