Wireless Sensor Network Continuous Plume Monitoring and Model Calibration: Proof of Concept in Intermediate-Scale Tank Test

The current practice for monitoring of subsurface plumes involves the collection of water samples from monitoring wells and laboratory analysis to determine concentrations. This data is used to make decisions for site management and in modeling. Cost and time constraints limit the number of samples and this approach becomes impractical for continuous monitoring of large, transient plumes. With the development of new sensor technologies and wireless sensor networks (WSNs), the potential exists to develop new and efficient subsurface data collection and monitoring methods. The goal is to automatically collect data from the sensors and wirelessly transmit the data to computer platforms where inversion codes and forward simulation models reside. This data can then be used to continuously monitor and update model parameters for the prediction of plume behavior. Many technological and operational challenges related to sensor placement and distribution, automation of real-time data collection, wireless communication, and modeling have to be overcome before the field implementation of complex plume monitoring systems. This preliminary proof of concept demonstration study assesses this technology using a physical aquifer test bed constructed in an intermediate scale tank. The test system includes a set of ten conductivity probes individually connected to wireless sensor boards (motes). The tank was packed using five well-characterized silica sands to represent a heterogeneous aquifer. Bromide tracer was continuously injected into a steady flow field and concentration at different points in the tank was measured with ten calibrated soil moisture/electrical conductivity sensors attached to six different motes. The motes in turn are connected to a computer for data analysis and coupled to models simulating flow and transport. The accuracy of the sensor-measured concentrations was tested against traditional grab samples analyzed using an ion chromatograph. Inverse modeling will be used to determine subsurface parameters needed for predictive modeling. This preliminary study is the starting step in the development of a more complex wireless sensing communication system to be used in field applications involving remediation design, performance assessment, risk analysis and exposure assessment.