Modelling potable-water consumption for the Service de l'Eau
Statistical analysis and ICI consumer segmentation for the Direction des réseaux d'eau, Ville de Montréal. Research intern, summer 2023.
Ville de Montréal · May–Sep 2023
title: 'Modelling potable-water consumption for the Service de l''Eau' summary: 'Statistical analysis and ICI consumer segmentation for the Direction des réseaux d''eau, Ville de Montréal. Research intern, summer 2023.' role: 'Research Intern' organization: 'Ville de Montréal' period: 'May–Sep 2023' stack: ['Python', 'R', 'Data analysis', 'Customer segmentation'] repoStatus: 'none'
Context
The Service de l'Eau, Direction des réseaux d'eau, needed a clearer picture of demand drivers across the City of Montréal's potable-water network. As a research intern on the team, I was given the brief to analyse historical consumption and identify the variables that mattered.
What I built
- Analysis of the city's consumption history to surface correlations and select the most relevant explanatory variables for downstream work.
- Consumer segmentation by ICI (industrial, commercial, institutional) usage category and geographic location, classifying subscriber profiles from historical data.
- Cross-referenced residential, ICI, and regulation-zone consumption alongside pipe-break records to surface the determinants of potable-water demand for the Water Networks Division.
What I took away
Working on a public-utility network gave me a habit I still use as a software engineer: take the data seriously, name what you can and can't conclude from it, and keep the analysis reproducible enough that someone else can pick up where you left off.
Stack
Python · R · Data analysis · Customer segmentation