Energy Consumption & Efficiency Analysis
Analyzed energy use, emissions, and efficiency across fire stations and office buildings. Applied regex-based data cleaning and statistical analysis to uncover operational insights.
Project Overview
This project analyzes energy consumption, emissions, and efficiency across various property types, focusing on fire stations and office buildings.
Methodology
1. Standardized column names and removed inconsistencies 2. Imputed missing values using median/mode 3. Applied regex for numeric and text cleaning 4. Cleaned postal codes and addresses
Regex Examples
extract_float_from_text extracts decimal values from numeric strings, while extract_postal_code_from_text formats inconsistent postal codes. Address strings were cleaned using clean_address_one to remove redundant city/province info.
Exploratory Data Analysis
- Summary stats and correlation analysis - Heatmaps and bar charts visualized energy use - T-tests compared Site EUI between property types
Outlier Detection
Used IQR and STD methods to detect and treat extreme values in energy consumption and emissions.
Challenges Faced
- Lack of monthly energy data limited seasonal insights - ENERGY STAR Score had too many nulls for comparison
Insights Gained
- Fire stations: 1.21 GJ/m², lower EUI than recreation facilities - Office buildings: 1.30 GJ/m², higher optimization potential - Energy use strongly correlated with emissions (0.76)
Visualization Highlights
