Maven Power Outage Challenge

Introduction

This Power BI dashboard project provides an analysis of power outages in the United States spanning the years 2003 to 2023. Leveraging data from the US Department of Energy, the dashboard aims to deliver insights into the frequency, causes, and impacts of power outages across various regions.

Data Cleaning: Before delving into the analysis, rigorous data cleaning was undertaken to address inconsistencies in the raw data. This involved standardizing date formats, resolving missing or erroneous entries, and ensuring data uniformity to facilitate accurate visualizations.

Data Quality Enhancements and Assumptions: In addressing data quality concerns and managing missing values, several modifications were applied to enhance the overall coherence of the dataset:

1. Handling Missing Event Start Time:
   • Instances where the event start time was absent were standardized to 00:00 hours. Additionally, a consolidation was performed by merging the date and time columns to streamline temporal representations.
2. NERC Region Standardization:
   • The NERC Region column underwent a rationalization process, condensing it into the current six regions based on qualifying states. This restructuring facilitates a more concise and regionally aligned analysis.
3. Area Column Transformation:
   • The 'Area' column was refined by redefining it as 'State' based on the first state name appearing in the respective row. This modification contributes to a more accurate geographical classification.
4. Type of Disturbance Categorization:
   • To streamline the dataset, the 'Type of Disturbance' underwent a regrouping process, resulting in a reduced set of seven distinct disturbance types. This simplification aids in clearer categorization and analysis.
5. Numeric Consideration for Demand Loss and Customers Affected:
   • For 'Demand Loss (MW)' and 'Number of Customers Affected', only rows with numeric values were considered, leaving others blank. This approach ensures a focus on quantifiable data points for more meaningful analysis.
6. Restoration Date and Time Handling:
   • Restoration Date and Time column were merged. Unknown values were converted to blanks.

Key Features:

1. Number of Incidents by Month:
   • Utilize a dynamic line chart to showcase the monthly distribution of power outages. This visual provides insights into seasonal patterns and helps identify trends over the years.
2. Average Demand Loss in Megawatts by Year:
   • Implement a trendline or line chart to represent the average demand loss in megawatts annually. This visualization offers a high-level overview of the magnitude of outages and their impact on the electrical grid.
3. Number of Incidents by NERC Region:
   • Present a bar chart highlighting the frequency of incidents across NERC regions. This visual enables stakeholders to identify regions with higher vulnerability and assess the effectiveness of region-specific resilience measures.
4. Bar Chart with Cause of Outage:
   • Implement a hierarchical bar chart allowing users to explore the causes of outages.
5. US Map Showing Incidents by State Using Bubbles:
   • Leverage a dynamic US map visualization with bubbles representing the number of incidents in each state. This map provides a clear spatial representation of outage hotspots, aiding in targeted intervention and resource allocation.

Key Takeaways: The months of August and February had the most incidents of power outage. Severe cold weather is the main cause in Feb, while August is mainly the Hurricane season in the South Eastern coast.

Conclusion:

In summary, the Power BI dashboard not only reveals historical patterns but also equips decision-makers with actionable insights to ensure a more reliable and robust power infrastructure for communities. Stakeholders can target investments and upgrades to address specific vulnerabilities, improving the overall reliability and resilience of the electrical infrastructure.