Emissions Data Quality & Transformation Pipeline

Problem

Emissions modelling relied on multiple raw data sources with inconsistent formats and varying data quality. Key inputs such as UK government CO₂ emission factors are updated annually, requiring a reliable system to track changes over time. Some values were based on fixed assumptions with no consistent way to document when they were last reviewed or where they originated. The lack of clear data lineage and auditability reduced trust in model outputs and made it difficult to trace errors in downstream calculations.

Key Work

Impact

Total Cost of Ownership (TCO) Data Model

Problem

Clients required a comprehensive Total Cost of Ownership (TCO) analysis to assess the financial feasibility of transitioning from ICE vehicles to electric vehicles. This involved integrating multiple cost components — vehicle costs, charging infrastructure, and more — into a single model. The challenge was to design a scalable data model that could accommodate new cost factors without significant schema changes, while maintaining fast runtime performance for user-driven recalculations.

Key Work

Impact

Example Visualisation

Total cost breakdown — Current (ICE) vs All EV Replacement scenario

Metric Current (ICE) Replacement (EV) Change
Total Cost £6,896,293.03 £8,823,426.51 +28%
Total Vehicle Costs £6,896,293.03 £7,913,426.51 +15%
CAPEX Vehicle Costs £1,382,500.00 £2,096,875.00 +52%
OPEX Vehicle Costs £5,513,793.03 £5,816,551.51 +5%
Total Location Costs £0.00 £910,000.00
CAPEX Location Costs £0.00 £850,500.00
OPEX Location Costs £0.00 £59,500.00
CO₂ Emissions 13,630.9t 5,531.7t −59%
Depot Charging vs Public Charging: Cost per kWh

Figure: Charging cost per kWh — depot (on-site) charging vs public charging stations.

The chart above compares the cost-effectiveness of investing in depot charging infrastructure versus relying on public charging stations. The analysis is based on energy consumption data derived from vehicle telematics, ensuring the comparison reflects real-world usage patterns.

DBT Optimisation

Problem

Product merges were taking a significant amount of time due to full dbt refreshes on large tables with high data volumes. This process often had to be run after hours, consuming considerable developer time and slowing down delivery workflows.

Key Work

Impact