American Society of Civil Engineers

Assessing the Impact of Population Distribution on Sensor Network Design

by Rakesh Bahadur, (Science Applications International Corporation, McLean, VA. E-mail:, William B. Samuels, (Science Applications International Corporation, McLean, VA. E-mail:, Robert Janke, (United States Environmental Protection Agency/National Homeland Security Research Center Cincinnati, OH. E-mail:, and Terra Baranowski Haxton, (United States Environmental Protection Agency/National Homeland Security Research Center Cincinnati, OH. E-mail:
Section: Water Quality, pp. 355-366, (doi:

     Access full text
     Purchase Subscription
     Permissions for Reuse  

Document type: Conference Proceeding Paper
Part of: Water Distribution Systems Analysis 2010
Abstract: This study examines the impact of dynamic population data on sensor network design. The primary driving force for managing and regulating distribution systems is to provide potable drinking water. Water distribution systems are inherently vulnerable to accidental or intentional contamination because of their distributed geography and numerous access points. The objective of a contamination warning system (CWS) is to minimize potential economic and public health impacts resulting from either an accidental release (such as a chemical spill) or a deliberate act of terrorism. Accidental or malicious introduction of a contaminant into water distribution systems could potentially have severe health effects on a population. Such intrusions can potentially be detected by deploying a number of sensing devices into the water distribution system. Optimal sensor placement needs to reflect variation in the timing of a contamination event and the distribution of the affected population. Population varies temporally (day versus night time) and spatially (different areas within city). The temporal population variation is not currently accounted for in sensor design studies. Previous studies of optimal sensor placement to support CWS design have considered only spatial variability using U.S. Census data or demand-based population models. This study included temporal variability by analyzing both day and night time population data. This was accomplished by using multiple population data sources such as LandScan and U.S. Census as input to the Threat Ensemble Vulnerability Assessment Sensor Placement Optimization Tool (TEVA-SPOT). This study showed that the temporal variation in population has a significant impact on sensor network design.

ASCE Subject Headings:
Probe instruments
Water distribution systems
Water pollution