Week 4: Data Crunch & First Visuals - Developing the Scoring Framework
This week was all about turning raw data into something actionable. I spent most of my time building Python scripts to process the CHP datasets and creating the first version of our scoring algorithm to rank thermal battery replacement opportunities.
Main Goals
- Clean and merge all CHP datasets into one unified database
- Build automated Python scripts for processing heat-to-power calculations
- Develop the initial scoring algorithm incorporating multiple factors
- Create first round of visualizations to validate our methodology
What I Completed
- Successfully cleaned and standardized all CHP datasets into a single MW-normalized master table covering prime mover type, fuel source, capacity, efficiency, and installation year
- Built Python automation scripts that calculate heat-to-power ratios across the entire dataset and generate facility age analyses
- Developed the initial scoring methodology that weighs H2P match, system age, cost factors, and fuel price volatility
- Created preliminary rankings for Michigan CHP facilities and validated the approach against known replacement scenarios
- Generated several key visualizations: cost trends for reciprocating engines over time, H2P distribution charts with scoring overlays, and sector-based capacity breakdowns
Key Insights The cleaned dataset is more comprehensive than I expected - it includes over 1,200 CHP installations across Michigan representing 2.8 GW of total installed capacity. About 34% of these existing installations fall within what appears to be the optimal H2P range for thermal battery replacement.
The scoring algorithm is showing promising results in early testing. Facilities that score above 75 points are achieving average cost savings of 18% when modeled for thermal battery replacement while maintaining equivalent thermal output.
Most interesting finding: reciprocating engines older than 15 years with H2P ratios between 2.2 and 3.1 consistently emerge as the highest-priority replacement targets. These combine favorable technical profiles with significant economic advantages.
Chart Generated Michigan CHP Replacement Scoring Distribution - A scatter plot showing H2P ratio versus facility age for all Michigan installations, with color-coded scoring that indicates replacement priority levels and projected economic savings potential.
Next Week Refine the scoring weights based on feedback, add more sophisticated economic modeling, and start preparing visualizations for the final presentation deliverable.
Week 4: From scattered data to strategic insights