“A low-cost ERT system from a new perspective: performance improvement and film thickness measurements under fluid electrical conductivity variations
Nome: ADRIANA MACHADO MALAFAIA DA MATA
Data de publicação: 18/08/2023
Banca:
Nome |
Papel |
|---|---|
| MARCIO FERREIRA MARTINS | Orientador |
Resumo: Electrical tomography is a non-intrusive technique connecting electrodes around a system. Through voltage measurements, the conductivity in the region inside the electrodes is estimated through the Inverse Problem (IP). By utilizing advanced techniques in the calculation of PI, obtaining tomography images quickly has become a streamlined process. These features enable safety control and multiphase flow measurement with low flow interference, in multiphase flow. Eidors is consolidated software in this technique and provides real and imaginary part results. Considering only the real part (resistive effects), we can call it Electrical Resistance Tomography (ERT). Electrical tomography depends on a conductive medium and can distinguish oil and gas
in water. Our line of research focused on identifying non-conductive objects in the continuous phase of water. Using a low-cost ERT acquisition system with varying conductivities, this thesis focused on data exploration and identification of variables with the most significant impact on reconstructed image quality. Most studies use tap water for testing, with conductivity up to 0.1 S/m. Oil extraction in deep waters brought the challenge of the presence of highly saline waters. Sea water can have high electrical conductivity, reaching up to 5.3 S/m, while formation water in oil extraction can have even higher values. The differential reconstruction is based on the measurement of the homogeneous tank compared with the tank with some non-conductive object.
Therefore, the differential reconstruction reduces the effect of the background conductivity since the solution is the same in both measurements. However, it is possible to observe that when increasing the solution conductivity, the image reconstruction is affected. Therefore, the study of the influence of conductivity is relevant. The electrical tomography technique has the challenge of unrelated data. Because it is highly sensitive, small changes can result in unpredictable changes in the reconstructed image. Therefore, measured data
may not be related to the applied tests. From the perspective of different conductivities, an analysis was proposed to assist in identifying critical variables. We identified the impact of the signal-to-noise ratio’s variance on the reconstructed image’s quality. In this way, it is possible to predict the final quality through the variance of the measurements and help in the adequate choice of techniques that propose the improvement of the images. Another challenge is that electrical tomography becomes unfeasible at high gas fractions. Alternatively, the film thickness calculation method allows using the same acquisition system. We then proposed improvements in modeling this technique with a dimensionless model to identify meaningful relationships.
In this way, it was possible to validate the results using the strategy of varying the conductivity of the solution.
