National Energy Technology Laboratory,
Keywords: sensors, corrosion
Summary:Electrochemical sensors are regularly used for the detection of chemical species in many industrial operations. The most common example is potentiometric measurement of pH, and oxygen quantity (e.g. the Clark electrode) and corrosion sensors also measure electrochemical processes. Systems using bulk aqueous streams provide a suitable electrolyte for most measurements. However, operation is currently limited in environments where water is distributed within a compressed gas or supercritical fluid. In such environments, the dispersed water does not provide adequate ion conductivity for conventional electrochemical sensors. Further, many sensor designs are not suitable for high-pressure operation. The development of solid state electrochemical sensors for non-aqueous environments will allow for improved monitoring of such streams for impurities and corrosion. National Energy Technology Laboratory and Pennsylvania State University prepared an initial design as part of their collaborative project for performing corrosion measurements in water-contaminated supercritical CO2 pipelines. These probes utilized a thin film of ion conducting polymer deposited across three electrodes, where moisture from the bulk CO2 phase allowed the polymer film to serve as the electrolyte. The study found that electrochemical corrosion measurements could be accomplished, although the quality of the deposited polymer film and its contact with the electrodes affected the measurements. A second project has improved upon this design and focused on its application to natural gas transmission pipelines. Commercial Nafion® membranes are used as the electrolyte to reduce the variability and contact issues with the polymer film. Electrode geometries have been adapted from test cells used for determining conductivity of polymer electrolyte membrane (PEM) fuel cells to provide reliable and repeatable contact. In this way, a five-electrode cell can be used to provide both corrosion measurements and accurate conductivity measurements of the membrane. As the conductivity of the membrane depends on the concentration of water in the gas stream, the results can be used to determine the water content of the stream. This is the determining factor for the corrosion risk of a gas pipeline, as corrosion generally does not occur without condensed water. While the conductivity of commercial membranes has been studied extensively for PEM fuel cells, almost all data is at ambient pressure. High-pressure conductivity measurements are required to calibrate the sensor for more aggressive environments. Initial results show promise that the current design can allow for low-cost solid-state sensors that can provide in-situ and real-time measurements of the water content and corrosivity in a natural gas pipeline. A second generation of the sensor could further improve reliability and decrease costs by depositing the electrodes and membrane onto a ceramic wafer. Once the platform for electrochemical measurements is established, it may be possible to incorporate additional electrodes, such as reference electrodes and ion selective electrodes, to increase the capabilities of the sensor. Such a sensor could be employed in a variety of gas and supercritical environments where the moisture content allows adequate conductivity of the membrane.