A Practical Statistical Method to Differentiate Inflow and Infiltration in Sanitary Sewer Systems
Robert B. Sowby, Daniel R. Jones
Abstract
Besides collecting legitimate wastewater, municipal sanitary sewer systems, even when separated from storm sewer systems, also receive unwanted inflow and infiltration (I/I) that adversely impact their sizing, economics, and operation. Although several statistical models have been proposed to quantify I/I, their advanced methods and intricate data requirements may be unrealistic for practitioners who need to investigate I/I for individual sewer systems. Here, a practical regression model of daily sewer flow is developed with discrete terms for sanitary flow (based on winter water use), groundwater infiltration (based on sinusoidal seasonality), direct inflow (based on same-day precipitation), and delayed inflow (based on multiday moving-average precipitation). The terms are intuitive and the model performs well with flow observations from a case study. The model can help practitioners separate I/I from other wastewater flows, customize measures to control I/I, and improve sewer system performance.