PowerPlots.jl: An Open Source Power Grid Visualization and Data Analysis Framework for Academic Research

Basic Information

• Paper ID: 2510.05063
• Title: PowerPlots.jl: An Open Source Power Grid Visualization and Data Analysis Framework for Academic Research
• Author: Noah Rhodes (Los Alamos National Laboratory)
• Classification: eess.SY cs.SY
• Publication Date: October 13, 2025 (arXiv v2)
• Paper Link: https://arxiv.org/abs/2510.05063

Abstract

Data visualization is crucial for understanding complex systems. The power grid is one of the world's most complex systems, and effective power grid research visualization software must satisfy three requirements: 1) ease of use; 2) support for unique data that may emerge in research; 3) ability to create customized figures for publication and presentation. However, no existing software simultaneously satisfies all three requirements. PowerPlots is an open-source power grid data visualization tool that meets these needs. Additionally, several tools created to support this software facilitate power grid data analysis by transforming data into graph topology or dataframe formats, making it more compatible with certain analyses. This paper uses PowerPlots to investigate several case studies involving power grid data exploration, demonstrating valuable insights that network visualization can provide and how it can be applied to research applications.

Research Background and Motivation

Problem Definition

The power grid, as one of the world's most complex engineering network systems, operates under constraints imposed by the physics of power flow. Researchers need visualization to:

1. Data Exploration: Rapidly and interactively visualize data, explore data patterns, and debug methods
2. Research Communication: Share specific information to convey research findings

Limitations of Existing Approaches

The author systematically analyzes the shortcomings of existing power grid visualization tools:

Limitations of Open-Source Tools:

• PyPSA, OpenDSS, PowerGridModel, PandaPower, etc., while providing basic network visualization, can only display conventional power research data and cannot handle unique data generated by novel research questions
• Lack capability to extend data formats and broadly customize figures

Limitations of Commercial Software:

• PowerWorld, PSSE, PLSF and other industrial software have limited flexibility and typically can only visualize data for which they were designed
• Not suitable or unable to adapt to research applications addressing novel power grid problems

Research Motivation

Power grid visualization holds significant value for researchers:

1. Helps identify unique behavioral patterns
2. Makes it easier to discover errors in computational methods
3. Develops intuition about power grid operation through pattern recognition

Core Contributions

1. Developed the PowerPlots.jl Framework: An open-source power grid visualization tool satisfying three principles of simplicity, flexibility, and customization
2. Created Supporting Data Structures: PowerModelsGraph and PowerModelsDataFrame to facilitate network analysis and data analysis
3. Provided Multiple Case Studies: Demonstrating the practical value of network visualization in power grid research
4. Implemented Broad Compatibility: Supporting PowerModels.jl and PowerModelsDistribution.jl data formats, capable of parsing MatPower, PSSE and other file formats

Methodology Details

Software Design Principles

PowerPlots.jl is built around three core design principles:

1. Simplicity: Provides intuitive default behavior for initial data exploration
2. Flexibility: Supports visualization of various customized power grid information
3. Customization: Enables high-level customization to create clear visualizations highlighting specific data

Plotting Pipeline Architecture

PowerPlots.jl employs the following four-step plotting pipeline:

1. Data Input: Uses the nested dictionary structure of PowerModels.jl as input
2. Coordinate Generation: Generates coordinate data for each component, creates a graph containing all component types, and applies graph layout algorithms
3. Data Transformation: Converts nested dictionary structures into a collection of dataframes, with each component type converted to an independent dataframe
4. Figure Rendering: Uses VegaLite.jl as the plotting engine, with each grid component rendered as a separate layer

Core Data Structures

PowerModelsGraph

An undirected graph structure containing all edges and vertices in the power grid:

mutable struct PowerModelsGraph
    graph::Graphs.SimpleDiGraph
    node_comp_map::Dict{Int,Tuple{String,String}}
    edge_comp_map::Dict{Graphs.AbstractEdge,Tuple{String, String}}
    edge_connector_map::Dict{Graphs.AbstractEdge, Tuple{String, String}}
end

• By default includes buses, generators, loads, and shunt reactors as nodes
• Branches, DC lines, switches, and transformers as edges
• Supports user-defined component inclusion

PowerModelsDataFrame

A dataframe structure facilitating data analysis:

mutable struct PowerModelsDataFrame
    metadata::DataFrames.DataFrame
    components::Dict{Symbol,DataFrames.DataFrame}
end

• Converts PowerModels.jl's nested dictionaries into tabular format
• Each grid component stored as an independent DataFrame
• Supports aggregated data queries across component types

Layout Algorithm Support

PowerPlots.jl supports multiple network layout algorithms, providing trade-offs between visual quality and computational speed:

• Kamada-Kawai: Default algorithm with best visual quality but slowest computation
• SFDP: Fast layout suitable for large networks
• Spring, Spectral: Balance between quality and speed

Experimental Setup

Test Networks

Primarily uses synthetic networks from the PGLib benchmark library:

• EPRI 39-node network
• IEEE 118-node network
• PEGASE 89-node network
• Large networks up to 1888 nodes
• RTS-GMLC system (with real geographic coordinates)

Functional Verification Scenarios

1. Network Structure Analysis: Comparing topological characteristics of different networks
2. Power Market Data Visualization: Locational Marginal Pricing (LMP) and transmission congestion
3. Multi-Network Time Series Data: Power grid restoration planning case
4. Distribution Network Support: Engineering model and mathematical model comparison
5. Interactive Data Exploration: Hover display of detailed information

Experimental Results

Network Topology Insights

Rapidly identifying network characteristics through visualization:

• PEGASE 89-node Network: High-density node clustering in central region with high connectivity
• IEEE 118-node Network: Typical transmission network structure lacking dense connections
• These observations align with network analysis conclusions in literature 32, confirming that certain PEGASE networks applied Kron reduction and similar processing

Node Degree Distribution Analysis

Analysis of node degree distribution across networks of different scales in PGLib shows:

• Small Networks (n<1000): Higher proportion of degree-1 and degree-2 nodes
• Medium Networks (1000<n<10000): Fewer degree-3 and degree-4 nodes
• Large Networks (n>10000): Maximum node degree can reach 41
• Small networks' maximum node degree of 15 may impact cascade failure analysis results

Power Market Data Visualization

Successfully demonstrated:

• Locational Marginal Prices: Price distribution shown through color encoding
• Transmission Congestion: Binding transmission limits identified in red
• Interactive Exploration: Hover to view detailed line parameters and power demands

Restoration Planning Case

Visualization of optimal restoration plan for 24-node network reveals:

• Repair strategy creates multiple power islands
• Inter-island connections occur in later restoration phases
• After connection, left power island provides significant power flow to right island

Distribution Network Support

Successfully supports two data models from PowerModelsDistribution.jl:

• Engineering Model: Represents actual grid components
• Mathematical Model: Equivalent network for optimization problems
• Supports multiple component types including switches, transformers, lines, buses, loads, and generators

Related Work

Comparison of Existing Open-Source Tools

• PyPSA: Supports standard power grid optimization models with basic network plotting
• OpenDSS: Distribution network simulator with graphical environment extension OpenDSS-G
• PowerGridModel: Supports power flow, state estimation, short-circuit analysis with data analysis extensions
• PandaPower: Emphasizes tabular data structure with basic network visualization support

Commercial Software Limitations

• PowerWorld: Industrial software with limited visualization capabilities
• PSSE, PLSF, ETAP, TARA: Certain degree of network data visualization but insufficient flexibility

Advantages of PowerPlots.jl

Compared to existing tools, PowerPlots.jl simultaneously possesses:

1. Simple and intuitive interface
2. Support for customized research data
3. Highly customizable publication-quality figure creation capability

Conclusions and Discussion

Main Conclusions

1. Fills Research Gap: PowerPlots.jl is the first power grid visualization tool simultaneously satisfying ease of use, data flexibility, and high customization
2. Facilitates Research Discovery: Case studies demonstrate that network visualization can rapidly identify data patterns and network characteristics
3. Supports Diverse Applications: Adopted by multiple research projects covering wildfire risk, carbon intensity metrics, restoration planning, and other domains
4. Provides Analysis Tools: PowerModelsGraph and PowerModelsDataFrame provide useful data structures for power grid data analysis

Technical Innovations

1. Unified Data Interface: Compatible with multiple file formats in the PowerModels.jl ecosystem
2. Flexible Component Support: Extensible to support new power grid component types
3. Multi-Level Customization: From simple parameters to full VegaLite.jl customization
4. Efficient Layout Algorithms: Multiple algorithm choices balancing quality and speed

Limitations

1. Julia Ecosystem Dependency: Requires users familiar with Julia programming language
2. Large Network Performance: Kamada-Kawai algorithm has longer computation time on very large networks
3. Learning Curve: Advanced customization requires understanding VegaLite.jl syntax
4. Real-Time Interaction Limitations: Primarily targets static visualization with limited real-time interactive features

Future Directions

1. Feature Extension: Continue adding new customization features based on community feedback
2. Cross-Domain Applications: Extend to other network systems such as natural gas, water, and telecommunications
3. Cross-Sector Analysis: Support cross-sector visualization analysis of critical infrastructure
4. Performance Optimization: Improve layout computation efficiency for large networks

In-Depth Evaluation

Strengths

1. Strong Practicality: Addresses actual visualization needs in power grid research, adopted by multiple research projects
2. Clear Design Philosophy: Three core principles (simplicity, flexibility, customization) guide sound software architecture
3. Solid Technical Implementation: Built on mature Julia ecosystem with well-designed data structures
4. Complete Documentation: Provides rich usage examples and detailed technical documentation
5. Open-Source Contribution: Provides valuable open-source tool for power systems research community

Weaknesses

1. Limited Evaluation: Lacks quantitative comparative analysis with other visualization tools
2. Insufficient User Research: No formal user experience evaluation conducted
3. Scalability Verification: Performance for ultra-large networks (>10,000 nodes) requires further validation
4. Cross-Platform Testing: Primarily tested in specific environments; cross-platform compatibility needs verification

Impact

1. Academic Contribution: Provides new tools and methods for power system visualization research
2. Practical Value: Adoption by multiple research projects demonstrates actual application value
3. Community Building: Promotes Julia application in power systems research
4. Standardization Advancement: Based on PowerModels.jl data format, facilitates tool standardization

Applicable Scenarios

1. Academic Research: Power systems researchers conducting data exploration and result presentation
2. Algorithm Development: Debugging and verification of new algorithms
3. Educational Applications: Visualization teaching in power systems courses
4. Engineering Consulting: Visualization needs in power grid planning and analysis projects

References

The paper cites 41 related references, primarily including:

• Power system visualization research 1
• Existing open-source tools: PyPSA 2, OpenDSS 3, PowerGridModel 4, PandaPower 6, etc.
• Julia ecosystem: PowerModels.jl 13, PowerModelsDistribution.jl 14, etc.
• Graph layout algorithms: Kamada-Kawai 33, Spring 35, SFDP 36, etc.
• Application cases: Wildfire risk 19,21-24, carbon intensity 20, network reconfiguration 25-29, etc.

Overall Assessment: This is a high-quality tools paper. PowerPlots.jl fills an important gap in power system visualization tools with good technical implementation and practical application value. While there is room for improvement in quantitative evaluation and user research, its open-source nature and existing application cases demonstrate its significant contribution to the research community.