A Practical Evaluation of Biostability, Chlorine Residual Loss, Material Sources and Impacts of Dead-End Flushing within Calgary’s Drinking Water Distribution System

Abstract

Drinking Water distribution system biostability is affected by water age, secondary disinfectant, nutrients, pH, temperature, dissolved oxygen and pipe material. Distribution water quality data were evaluated for the City of Calgary and chlorine decay rates were determined experimentally. A C:N:P ratio analysis revealed total organic carbon (TOC) as the limiting nutrient for biological regrowth followed closely by phosphorus. Glenmore Water Treatment Plant treated water is more biosustaining when compared to Health Canada’s dissolved organic carbon guideline of 1.8 mg/L. Correlation analysis between cellular adenosine tri-phosphate (cATP) (as biological regrowth) and key factors demonstrated a weak, but significant statistical relationship between cATP and free chlorine. Weak non-significant relationships were found for cATP:TOC and cATP:pH. Climate change impacts, specifically warmer air temperatures impacting both ground and water temperatures, increased precipitation volatility and impacts of wildfires were considered. Chlorine decay scenarios were evaluated under warmer water conditions and elevated TOC. While biostability is not a current concern for the Calgary distribution system, there is a risk of increased biological regrowth and chlorine residual loss should a 1 mg/L TOC and ~2˚C water temperature increase occur. To explore a materials mass balance discrepancy determined by a desktop analysis of historical treated and distribution water quality data, dead-end flushing and water quality sampling was completed at polyvinyl chloride (PVC) and iron pipe sites across Calgary. Metals, nutrients and biological regrowth were statistically compared to pipe characteristics and flushing metrics were developed. Iron pipe dead-ends >250m had a higher risk of chlorine residual loss, material accumulation and biological regrowth. PVC pipe sediments were dominated by aluminum and demonstrated greater biostability. At 1 m/s (3.3 ft/s) only 4 pipe volumes or 15 minutes was needed to clean >85% of the material from the dead-end mains and would conserve significant water volumes. Analyses indicate corrosion is the dominant material source while treated water loading and legacy deposits are secondary material sources. This project adds to the limited body of work concerning dead-end flushing and demonstrates how dead-end flushing can be used to better understand distribution water quality.