Evaluation of Sewer System Rehabilitation via Flow and Storm Data Analysis

Using rainfall and metered in-pipe wastewater flow time series data to assess the integrity or a rehabilitation effect of a sewer basin involves two analytical stages: Stage 1) estimating the total rainfall (x) and the amount of inflow and infiltration (Y ) for each of the metered rain events; and Stage 2) modeling the (x,Y ) points from two comparative data series (either of two neighboring basins or pre- versus post-rehabilitation on a same basin), and making model-based inferences on whether and how much of a reduction in inflow and infiltration is achieved under a hypothesized n-year storm. The analysis is nontrivial, and serious technical issues exist at each of the two stages. After a brief discussion on the issues of Stage 1 analysis, this paper focuses on several reasonable models for Stage 2 analysis. The main issue of the discussion is on the extrapolating nature of such models for severe n-year storms which are often out of observed range in a short-term data series typically lasting only for several months.