Sensitivity of DC Network Representation for GIC Analysis

Aniruddh Mishra, Arthur K. Barnes, Jose E. Tabarez, Adam Mate · 2025

Read Sensitivity of DC Network Representation for GIC Analysis by Aniruddh Mishra, Arthur K. Barnes, Jose E. Tabarez, Adam Mate (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-23016.txt.

Introduction

Aniruddh Mishra, Arthur Barnes, Jose Tabarez, and Adam Mate wrote this paper in two thousand twenty-five. It tests how touchy power loss sums are to the way a grid is drawn in direct current form, for studies of geomagnetically induced currents.

Storms in space drive such currents into grids, where they can cook transformers. Guards must be placed at the best spots, and that search needs a DC map of the AC lines. Choices in that map, such as how guard units are drawn, swing the loss sums by much. On a sample network the authors set the variants side by side. The work warns planners to mind the map before they trust the plan.

Catalog record

CREATORAniruddh Mishra, Arthur K. Barnes, Jose E. Tabarez, Adam Mate
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-23016

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Sensitivity of DC Network Representation for GIC Analysis Thanks: Approved for public release; distribution is unlimited. LA-UR-24-31515 Thanks: 979-8-3315-4112-5/25/$31.00 ©2025 IEEE Aniruddh Mishra, Arthur K. Barnes, Jose E. Tabarez, Adam Mate Abstract Geomagnetic disturbances are a threat to the reliability and security of our national critical energy infrastructures. These events specifically result in geomagnetically induced currents, which can cause damage to transformers due to magnetic saturation. In order to mitigate these effects, blocker devices must be placed in optimal locations. Finding this placement requires a dc representation of the ac transmission lines, which this paper discusses. Different decisions in this process, including the method of representing the blocking devices, result in significant variations to the power loss calculations. To analyze these effects, we conclude the paper by comparing the losses on a sample network with different modeling implementations. I Introduction Geomagnetic disturbances (GMDs) adversely affect energy infrastructures; they result in geomagnetically induced currents (GICs) throughout the transmission network, which enter the electric grid at low resistance paths to ground, particularly at neutral wires of transformers. Though GMDs can result in a range of contingencies in the grid, the primary impact we will be investigating is the effect on high-voltage power transformers. Since most transformers are rated for only a few additional amps while performing in a linear operating region, the addition of GICs into the network could result in extended power outages [ 1 , 2 , 3 ] . Modeling GIC flow consists of two steps: (1) the generation of a dc network from a given ac model, as GICs are pseudo-dc phenomenon; and (2) utilizing a simulation and analysis software to find solutions for power flows and mitigate GMD impact. It is possible to model GICs with an EMTP-type simulation [ 4 ] , but it results in high computational complexity and an inability to implement optimization for mitigation contexts. The method of dc equivalent circuits [ 5 ] is more computationally efficient, is feasible for use in an optimization context [ 6 ] , and can be coupled to a positive sequence equivalent circuit with a transformer saturation model [ 7 , 8 ] . A summary of this method for typical power systems components is provided in [ 1 ] . GIC blockers are designed to mitigate dc current without affecting the ac network. The operating concepts are similar to that of a capacitor with the added requirement of not introducing any voltage transients, such as ferroresonance [ 9 ] . There are several techniques to optimize placements with a given blocker type. They range from extensively searching every combination of open and closed lines in the dc network to implementing a mixed-integer optimization problem [ 6 , 10 ] . Overall, the problem can be stated as finding the minimum total power loss with the number of blockers, B , less than or equal to a maximum quantity, N_{B} . Previous work has studied the sensitivity of GIC flow with respect to network parameters [ 11 ] , specifically the impact of line resistance and grounding resistance on GIC magnitudes. However, the sensitivity of different dc network representations on the blocker placement problem has not been analyzed. Additionally, the partial blocker placement problem is a challenging and active area of research as placing blockers at one substation can increase GIC-induced transformer reactive power loss ( Q_{loss} ) at other substations [ 6 ] . A better understanding of the trade-offs between different blocker representations will help researchers develop scalable algorithms for this problem. This paper will examine sensitivity to different modeling decisions of blockers. Specifically, in this paper, we will take into account the differences in three main dc-current blocking devices: line series capacitors, transformer neutral blockers, and substation ground approximation of neutral blockers. This will extend to a discussion of how different blocker placement representations vary results. Finally, we will contribute to the analysis of uniform and non-uniform electric field models and their impact on power loss output. This work relies on PowerModelsGMD.jl (abbrev. PMsGMD) 1 1 1 https://github.com/lanl-ansi/PowerModelsGMD.jl for both dc network generation and analysis [ 12 ] . For the case of transformer neutral blockers, we relied on PowerWorld Simulator for results, as PMsGMD does not currently model these devices exactly and relies on the substation ground blocking approximation. The general organization of the paper is as follows: Section II discusses generation of dc networks. Section III covers the calculations of the induced voltages on transmission lines. Section IV describes the conditions that lead the assumption of implicit components. Section V describes the modeling assumptions associated with different blocking options. Section VI describes the experimental procedure and case study system. Finally, sections VII and VIII present results from the case study system and overall conclusions, respectively. II DC Network Generation II-A Constructing GMD Buses The main difference between the ac bus table and dc bus (GMD bus) table is the addition of substation grounding nodes, which do not have corresponding components in the ac model. Since the additional branches in the dc network are the decomposition of transformer configurations, an extra bus at the ground of substations is required to represent all of the dc branches (GMD branches). TABLE I: Mappings Between Transformer Terminals and Indices of Buses and Branches Terminal ac dc dc Bus Bus Branch High-Side \eta(k) h(k) H(k) Low-Side \lambda(k) l(k) L(k) Tertiary \tau(k) t(k) T(k) Series s(k) S(k) Common c(k) C(k) Ground g(k) II-B Constructing GMD Branches GMD branches represent any path that GICs flow through. There are two main types of GMD branches: (1) the resistive components of ac transmission lines; and (2) windings of transformers that are connected to the substation ground. Additionally, implicit GMD branches are also generated from bypassed series capacitors as discussed in section IV-C . II-B 1 Lines The resistance of transmission lines in Ohms is calculated as follows: R_{si}=R_{pu}^{1}\left(\frac{kV_{ll}^{2}}{3\text{{MVA}}_{3\phi}}\right) (1) In the above R_{si} is the common-mode dc resistance of the branch, R_{pu}^{1} is the positive-sequence resistance of the branch in per-unit, kV_{ll} is the rated line-line voltage in kV , and \textit{MVA}_{3\phi} is the three-phase base power in MVA. II-B 2 Transformers Y_{g} - Y_{g} : Both the winding configurations of the Y_{g} - Y_{g} transformer have a connection to the ground of the substation. The dc current in these transformers can flow through the windings to ground, which completes a loop for GIC flow. To represent these extra paths for dc current, GMD branches are added on viable transformer windings. Particularly, any winding that is connected to ground has a GMD branch associated with it. The resistances of these branches are given as a third of the real part of the impedance of the windings to account for parallel phases. In addition, a GMD bus is added at the substation grounding node, as discussed in Section II-A . This is the substation grounding node, labeled \textsl{g}(k) , in Fig. 1 [ 1 , 11 ] . Fig. 1: DC equivalent circuit of a Y_{g} - Y_{g} transformer [ 12 ] \Delta - Y_{g} : These transformers do not have GMD branches from both buses to ground. Since the \Delta windings do not have a direct connection to two GMD

Read the complete text