Automation of Water Distribution System Model Build-up
Väyrynen, Janne (2021)
Väyrynen, Janne
2021
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2021060814984
https://urn.fi/URN:NBN:fi:amk-2021060814984
Tiivistelmä
International Water Association (IWA) has listed smart and intelligent network technologies as one of the major trends in the water industry in the next decade (IWA, 2020). Hydraulic models help water utilities become more sustainable by reducing leaks, optimizing the network investments, and the system's energy use. This thesis, commissioned by Fluidit Ltd., had the objectives of documenting the current Water Distribution System (WDS) model build-up used by Fluidit Ltd. and automating the process by implementing scripts and pseudo-codes. The aim was to improve network processing, which has usually been the most time-consuming part when constructing hydraulic network models. The validation methods created in this thesis aim at easily accessible, better-quality hydraulic models at a global scale.
In this thesis, the methods for documenting and improving the model build-up were gradually improved. At first, the current model build-up process was documented. Manual processing phases were replaced step by step with scripts and the pseudo-codes implemented in this thesis. The processes include validating network connectivity and the essential pipe attributes in the WDS models. Finally, the two hydraulic model building methods, manual and import methods, were compared in terms of quality and speed.
The improved model build-up process was documented in this thesis, and it will be used when training modelers using Fluidit software. The modeling process includes the requirements of input data and how to deliver the data to the model. The network data was put under validation using scripts and the pseudo-codes created in this thesis, and a method for the automatic WDS network model update was introduced. Implementing the automatic build-up tool was not fully completed during the thesis; thus, the performance could not be thoroughly evaluated. The comparison of the manual building with a semi-automatic import process showed that the manual build-up could be efficient when modeling small networks in terms of speed, but importing the network is beneficial in terms of model quality, improving the quality of the original network data, and reducing data loss when creating a hydraulic model. The developed automatic validation method has shown promising results with actual network data. Visual, map-based network validation reports made it easier to correct the network in network information system (NIS), as the flaws in data became more visible. For example, a test case using data from a Finnish water utility showed that the topological connectivity of the modeled network had significant improvement just after two iterations of network validation. The data validators and the model building processes will be further improved after this thesis, adding implementations in sewer, storm, and district energy networks. Some input data sources that limit fully utilizing the hydraulic model build-up in WDS networks are listed in Appendix 1.
In this thesis, the methods for documenting and improving the model build-up were gradually improved. At first, the current model build-up process was documented. Manual processing phases were replaced step by step with scripts and the pseudo-codes implemented in this thesis. The processes include validating network connectivity and the essential pipe attributes in the WDS models. Finally, the two hydraulic model building methods, manual and import methods, were compared in terms of quality and speed.
The improved model build-up process was documented in this thesis, and it will be used when training modelers using Fluidit software. The modeling process includes the requirements of input data and how to deliver the data to the model. The network data was put under validation using scripts and the pseudo-codes created in this thesis, and a method for the automatic WDS network model update was introduced. Implementing the automatic build-up tool was not fully completed during the thesis; thus, the performance could not be thoroughly evaluated. The comparison of the manual building with a semi-automatic import process showed that the manual build-up could be efficient when modeling small networks in terms of speed, but importing the network is beneficial in terms of model quality, improving the quality of the original network data, and reducing data loss when creating a hydraulic model. The developed automatic validation method has shown promising results with actual network data. Visual, map-based network validation reports made it easier to correct the network in network information system (NIS), as the flaws in data became more visible. For example, a test case using data from a Finnish water utility showed that the topological connectivity of the modeled network had significant improvement just after two iterations of network validation. The data validators and the model building processes will be further improved after this thesis, adding implementations in sewer, storm, and district energy networks. Some input data sources that limit fully utilizing the hydraulic model build-up in WDS networks are listed in Appendix 1.