| dc.description.abstract | This thesis delves into the complex issue of sand production in the oil and gas industry, with a focus on the role and performance of Wire Wrapped Screen and 20 Mesh sand control systems. Recognizing that sand production is a significant challenge in well operations and is not yet API standardized, this research conducts an experimental study on these two types of screens, each loaded with various sand pack sizes.
The experiments illuminate key performance aspects of the screens, such as pressure build-up phenomena, sand production correlation with sand pack size, effective permeability, and flow rate fluctuations. The study reveals that pressure build-up varies with sand pack volume and screen design, and sand production does not show a clear correlation with sand pack size in the two different screen designs and sizes. The flow rate fluctuation is observed to be higher through the 20 Mesh, which could be due to higher bridge instability in this screen.
Beyond the experimental investigation, this thesis also applies two machine learning models, Random Forest (RF) and Artificial Neural Network (ANN), to predict sand production. The models demonstrate promising accuracy, with R2 values of 0.926 and 0.988, respectively.
This research paves the way for further studies in this field, suggesting future investigations on rig modification, sand production tests with different particle size distributions, and horizontal radial flow rig tests. The author concludes that the advancements in understanding sand control screens made in this thesis, along with the suggested future work, will contribute significantly to tackling the problem of sand production in oil and gas wells. | |