Chronic Source Water Quality Exceedances in the Grand River Watershed: Implications for AI Data Center Cooling Infrastructure in Ontario, Canada

This study analyzes 7,630,482 measurements from the Ontario Provincial Water Quality Monitoring Network (PWQMN) spanning 1964 to 2024, encompassing 438 monitoring stations across Ontario. We find that 41 of 42 monitored Grand River stations (97.6%) maintain mean conductivity and/or turbidity values exceeding Ontario Drinking Water Quality Standards on a chronic, non-episodic basis. Conductivity across the watershed averages 717 µS/cm against a 700 µS/cm operational guideline, with individual stations recording means up to 1,522 µS/cm and peak values of 10,069 µS/cm. Turbidity averages 23 FNU across 95% of stations against a 4.0 FNU standard. These findings are independently corroborated by City of Brantford municipal water quality records (2021–2026). We conclude that AI data center cooling towers drawing from Grand River municipal supply face elevated and measurable risk of heat exchanger scaling, corrosion, and biofouling. The Presignal Subtraction Framework is proposed as a continuous monitoring layer for early warning of chemistry drift before operational failures occur.