PipeNetGen: A methodology for generating synthetic water distribution network models

Area of Science:

• Civil Engineering
• Environmental Engineering
• Water Resource Management

Background:

• Water distribution networks (WDNs) are essential for delivering water but modeling them is resource-intensive.
• Under-resourced utilities often lack the capacity to develop accurate WDN models.
• Effective WDN models are crucial for infrastructure design, operation, and strategic decision-making.

Purpose of the Study:

• To develop an automated methodology for generating synthetic Water Distribution Network (WDN) models.
• To overcome the resource limitations faced by water utilities in creating WDN models.
• To provide a tool for improved water infrastructure management and research.

Main Methods:

• Developed the Pipe Network Generator (PipeNetGen) methodology.
• Integrated graph theory, Mixed-Integer Linear Programming (MILP), clustering algorithms, and hydraulic simulation.
• Incorporated a fire flow method for network reliability assessment.

Main Results:

• PipeNetGen successfully generated a synthetic WDN model for Lakewood, California.
• The synthetic model showed comparable results to an engineered model in terms of pipe sizes, topology, hydraulics, and demand.
• The methodology demonstrates feasibility for creating WDN models from open-source or private data.

Conclusions:

• PipeNetGen offers an efficient and valuable resource for water infrastructure management.
• The methodology can assist utilities with limited resources in developing crucial WDN models.
• This approach supports advancements in WDN research and decision-making processes.