Document Type
Presentation
Publication Date
8-1-2014
Abstract
Currently, achieving appropriate operative performance of water infrastructure has become a high priority in urbanized areas. Particularly, providing reliable sewerage service is central for human well-being and its development (Kleidorfer, et al. 2013). Having that wastewater system management is an increasingly complex task due to a number of hardly predictable factors (e.g. deterioration of system components and climate variability), recent research efforts have been focusing on developing methods to identify optimum proactive rehabilitation and maintenance strategies, some of which are based on the identification of the sewerage structures in most need of attention. To meet such a goal, different forecast failure models for urban water infrastructure have been recently developed. These models are able to assess the future behavior of water supply and sewer system structures. This study presents the comparison of two different failure statistical packages for urban water systems: (a) The FAIL software that calculates failure predictions based on two alternative stochastic processes, the single-variate Poisson process and the Linear Extended Yule process (LEYP) (see Martins et al., 2013) and (b) The SIMA software that, trough out a series of statistical tests, selects a failure model that is based either on an homogeneous Poisson process (HPP), a renewal process or a non-homogeneous Poisson process (NHPP), which allows changes of trend in the failure intensity (see Rodríguez, et al. 2012). Those different statistical models are applied to two contrasting urban wastewater systems: Bogotá (Colombia, 7.5 million inhabitants) and Oeiras e Amadora (Portugal, 10.000 inhabitants). Customer complaints and failure databases were gathered in order to analyze two different types of sewer failures named sediment-related blockages and structural failures. Multiple analyses are carried out in order to assess the impact of sewer system characteristics, system complexity, spatial resolution and data availability onto models forecasting efficiency.
Comments
Session R52, Sanitary Sewer Networks