Principal component analysis for the process of opening underground caverns in salt rocks

• Ano de defesa: 2022
• Autor(a) principal: SILVA, Rayra Amorim
• Orientador(a): GUIMARÃES, Leonardo José do Nascimento
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Pernambuco
• Programa de Pós-Graduação: Programa de Pos Graduacao em Engenharia Civil
• País: Brasil
• Palavras-chave em Português: Engenharia civil; Caverna; Rocha salina; Variáveis de processo; Simulações numéricas; Análise estatística
• Link de acesso: https://repositorio.ufpe.br/handle/123456789/46385

The constant dynamism within the oil industry associated with the need for new technologies in terms of production and disposal of products were fundamental for the increase of studies about the use of underground caverns in salt rocks as an alternative for the storage of petroleum products. Salt rock is particularly useful for storage because of its low cost, low permeability, and its healing potential when compared to other rocks, including granite, mud, and basalt. The opening process and subsequent development of these cavities are complex activities and the variables involved in the process play a crucial role during the entire operation. In this sense, the present work aims to identify, through the PCA (Principal Component Analysis) statistical tool, the variables that most influence the process of opening a salt cavern by dissolution. For this, numerical simulations of the dissolution mining process for opening a cavern under typical conditions of water injection into a salt rock using the software SALGAS were developed considering different methods of saline water circulation, after that, the variables injection temperature, injection rate, radius, volume, pump power, cumulative energy, tubing loss, produced brine, pump pressure, injection pressure, and salt dissolution factor were interpreted using the multivariate statistical tool through software PAST. For the simulations generated, the results with the statistical tool were satisfactory, it was found that the brine injection rate contributes significantly to the process, in terms of x-axis, directly influencing the behavior of other variables, the temperature have a great importance to the y-axis. Regarding the total variability of the data, more than 97% of these could be represented in terms of the first two components for both scenarios studied.