An approach to leak detection using wireless sensor networks at carbon sequestration sites

This paper concerns the problem of detecting leaks at carbon sequestration sites through the monitoring of CO2 levels using a wireless sensor network (WSN). By applying a basic linear dynamic model for an advection–diffusion process, a model-based detection strategy called the Iterative Partial Sequential Probability Ratio Test (IPSPRT) can be employed to detect and localize multiple leaks. A 3-D CO2 transport model is employed to provide a proof of concept simulated evaluation of the IPSPRT against a windowed-average approach in terms of time-to-decision vs. probability of false alarm and probability of a missed alarm. Despite using only a basic linear dynamic model, the IPSPRT is shown to achieve exponentially increasing better time-to-decision than the windowed-average approach as the probability of false alarm and probability of a missed alarm are decreased. A test bed implementation is described and employed to evaluate the performance of the IPSPRT in the presence of common WSN errors.

► We introduce the IPSPRT to perform CO2 leak detection using stochastic methods.
► A proof-of-concept evaluation of the IPSPRT is provided through simulations.
► The IPSPRT is implemented in a wireless sensor network test bed.
► The IPSPRT is evaluated in the presence of common wireless sensor network errors.