- Library Home /
- Search Collections /
- Open Collections /
- Browse Collections /
- UBC Theses and Dissertations /
- Advanced deterioration models for wastewater inspection...
Open Collections
UBC Theses and Dissertations
UBC Theses and Dissertations
Advanced deterioration models for wastewater inspection prioritization Balekelayi, Ngandu
Abstract
Developing an accurate deterioration model for sewer pipes is the basic step to a successful wastewater infrastructure management. However, it is difficult to predict exactly the change in the condition of sewer pipes since they are buried, and several factors influence their condition changes. These factors are collected through inspection and the deterioration models are built to understand the way they affect the condition of sewer pipes. However, only few pipes are inspected and not all the needed factors are collected. Thus, models are made from uncertain and incomplete data resulting in simplified approaches for quick and easy interpretation of the results. In this research, two semi-parametric models are developed to predict the deterioration of sewer pipes. The first model predicts the expected mean deterioration score of sewer pipes and the second gives the complete distribution of the output response through the analysis of its quantiles. In both models, variables are grouped into categorical where an ordinary least square estimation of parameters is applied and continuous variables where nonlinear smooth functions P-splines are applied. To account for unobserved covariates, models are enhanced through the inclusion of geospatial location of pipes as surrogate covariate that is estimated through a Gaussian Markov Random Field approach. The application of these models to the sewer inspection data of the city of Calgary shows a good agreement between the predictions and the observations. The estimation of the risk of pipes failure passes through the identification of critical areas in the network and their rankings, a metric resulting from the aggregation of multiple graph theoretic centrality measures is calculated. Two frameworks are proposed including the Bayesian approach when the hydraulic data are available to estimate the baseline metric and the Ordered Weighting Averaging that considers the decision maker’s attitude. Finally, the value of the information provided by the deterioration models is used to orient future inspections in the network. From the deterioration models’ information and errors, the value of information is calculated for each pipe in the network. The obtained values are projected on a GIS map for different horizons to help decision makers orient their priorities.
Item Metadata
Title |
Advanced deterioration models for wastewater inspection prioritization
|
Creator | |
Publisher |
University of British Columbia
|
Date Issued |
2019
|
Description |
Developing an accurate deterioration model for sewer pipes is the basic step to a successful wastewater infrastructure management. However,
it is difficult to predict exactly the change in the condition of sewer pipes
since they are buried, and several factors influence their condition changes.
These factors are collected through inspection and the deterioration models
are built to understand the way they affect the condition of sewer pipes.
However, only few pipes are inspected and not all the needed factors are
collected. Thus, models are made from uncertain and incomplete data resulting in simplified approaches for quick and easy interpretation of the
results. In this research, two semi-parametric models are developed to predict the deterioration of sewer pipes. The first model predicts the expected
mean deterioration score of sewer pipes and the second gives the complete
distribution of the output response through the analysis of its quantiles. In
both models, variables are grouped into categorical where an ordinary least
square estimation of parameters is applied and continuous variables where
nonlinear smooth functions P-splines are applied. To account for unobserved
covariates, models are enhanced through the inclusion of geospatial location
of pipes as surrogate covariate that is estimated through a Gaussian Markov
Random Field approach. The application of these models to the sewer inspection data of the city of Calgary shows a good agreement between the
predictions and the observations. The estimation of the risk of pipes failure
passes through the identification of critical areas in the network and their
rankings, a metric resulting from the aggregation of multiple graph theoretic
centrality measures is calculated. Two frameworks are proposed including
the Bayesian approach when the hydraulic data are available to estimate
the baseline metric and the Ordered Weighting Averaging that considers the
decision maker’s attitude. Finally, the value of the information provided by
the deterioration models is used to orient future inspections in the network.
From the deterioration models’ information and errors, the value of information is calculated for each pipe in the network. The obtained values are
projected on a GIS map for different horizons to help decision makers orient
their priorities.
|
Genre | |
Type | |
Language |
eng
|
Date Available |
2019-06-06
|
Provider |
Vancouver : University of British Columbia Library
|
Rights |
Attribution-NonCommercial-NoDerivatives 4.0 International
|
DOI |
10.14288/1.0379341
|
URI | |
Degree | |
Program | |
Affiliation | |
Degree Grantor |
University of British Columbia
|
Graduation Date |
2019-09
|
Campus | |
Scholarly Level |
Graduate
|
Rights URI | |
Aggregated Source Repository |
DSpace
|
Item Media
Item Citations and Data
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International