Markov chain modulated Poisson process to stimulate the number of blockages in sewer networks

Abstract

Blockage failure is the most common type of operational failure in sewer networks that can cause loss of service and flooding, which can result in environmental pollution, health risks, property damage, and traffic disruption. There are currently very few blockage prediction models that are either deterministic in nature or depend only on static factors in predicting blockages. This study aims to overcome these drawbacks in the current blockage prediction models and proposes a methodology that aims to predict the expected number of blockages in sewer networks as a function of pipe conditions (dynamic variable) and pipe physical attributes (static variables) using a Markov chain modulated Poisson process modeling framework. The framework is applied to case study sewer network in the city of Sahab, Jordan, that contains information about the pipes’ physical attributes and their current condition. Bayesian analysis is then performed to evaluate the proposed framework.