Detalhes bibliográficos
| Ano de defesa: |
2011 |
| Autor(a) principal: |
Souza, Anderson Dantas de
 |
| Orientador(a): |
Santana, Pedro Leite de
|
| Banca de defesa: |
Não Informado pela instituição |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal de Sergipe
|
| Programa de Pós-Graduação: |
Pós-Graduação em Engenharia Química
|
| Departamento: |
Não Informado pela instituição
|
| País: |
BR
|
| Palavras-chave em Português: |
|
| Palavras-chave em Inglês: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
https://ri.ufs.br/handle/riufs/5049
|
Resumo: |
The multiphase flow is a subject that encloses a vast field of knowledge and applications, different technological contexts, different scales, and is target of relatively recent studies. As basic examples there are industrial transport processes as water-vapor, fluidized beds and transport of oil. It can be said that, amongst these systems, the oil transport is presented as classic example of the multiphase flow, therefore can be observed on it all the complexities: flow that involves all the possible phases, that is, solid-liquid-liquid-gas, for particles in suspension (silicon, resins and asphaltenes, metallic composites and salts), oil (liquid hydrocarbons), water and gas (gaseous hydrocarbons), respectively. However, it must be detached that the multiphase flow usually is dealt with some assumptions. The knowledge of the multiphase flow characteristics also is basic for the equipment development of fluids properties measurement on-line, as well as measurement of outflow and pressure, variable of basic interest for the management of reservoirs, quantitative transference control of fluids produced between producer and purchaser, management control of emptyings, fiscalization, amongst others. This work presents a methodology with the use of artificial intelligence techniques, specifically those basing on Artificial Neural Network - ANN's, to predict pressure drop and gradient pressure in multiphase flow, assuming the Black Oil physical model, for different gaseous phase mass fractions in the start of the flow, taking in account properties of the flow, such as viscosities of the individual phases and the mixture, specific mass and speeds of the phases, emphasizing itself flow situations that occur in the oil industry. For the definition of the ANN's architectures and training algorithms it was used data gotten with the deterministic models solutions. It was used, specifically, the deterministic homogeneous and separated flow models. The simulations gotten with the ANN s used had been compared with those solutions gotten with the deterministic models, verifying itself that the used methodology presents satisfactory precision and simplicity of use, compatible with the necessities of the oil industry, being able the boarding to be extended to the situations where operational data are available. |
