Desenvolvimento de dispositivos a base de nanomateriais de carbono para sensoriamento de CO2

Abstract

One of the technological and industrial challenges generated after the pre-salt discovery is the detection of carbon dioxide (CO2) present in natural gas (NG). This contaminant (CO2) is a possible cause of internal corrosion in oil and natural gas installations. Moreover, for distribution in the national territory, is necessary to guarantee the quality of the natural gas processed with minimum requirements of this contaminant, insomuch that they meet the technical specifications following with the legislation. However, it is difficult to develop sensors for detecting CO2, as is a highly corrosive gas, that needs to be detected in a high-pressure environment. Therewith, is necessary to develop a sensor with a material resistant to the corrosive environment, and which presented uncomplicated integration into the production chain. Because of this motivation, this project aimed to meet Petrobras demand for sensors with a compact size, quick and selective response for the detection of the contaminant CO2 in natural gas. Recent studies indicate the great potential of carbon nanomaterials for the application as sensors because of their electrical, optical, and structural properties. The activities developed during this project involve the production and the electrical characterization of the nanodevice in the atmosphere of CO2. In this work, manufacturing nanodevices using functionalized nanotube ink and the electrical characterization system developed only for simulation in a relevant environment will be presented. We will show from the first detections of CO2 in an argon environment, the detection of CO2 in a methane environment (CH4), and finally the detection of CO2 in a natural gas environment, which is approximately from a real application. Results obtained from measurements of the sensor response in real-time, during the detection of CO2 in a NG environment, tested to be a sensor capable of detecting a wide range of concentrations (0 - 100%) with a time of 10min between measurements, at room temperature and atmospheric pressure. Through several tests, we demonstrate that the sensor remains stable and with a lifetime of over 8 months. We will also present the test carried out with over 2000 measurements of concentrations of CO2 applied randomly, getting a sensor response curves capable of distinguishing the applied concentrations without sensor saturation. An approach using Machine Learning will be presented to perform the calibration and estimated of the concentrations through the measurements performed by the sensor. To complete the work, the proposals for the detection mechanism of CO2 will be valid. Therefore, this work presents a CO2 sensor with the possibility of scalable production and integration into the production chain for monitoring the quality of natural gas.