Modelo de regressão Weibull para estudar dados de falha de equipamentos de sub-superfície em poços petrolíferos

Abstract

Este trabalho apresenta um estudo de confiabilidade em dados relativos ao tempo de vida de poços petrolíferos terrestres da Petrobras, produtores de óleo na Bacia Potiguar (RN/CE). O objetivo do estudo foi, com base em um conjunto de dados sobre ocorrências de falhas, verificar a existência do relacionamento entre o tempo de vida dos poços e algumas de suas características, como método de elevação, nível de produção, BSW (Basic Sediments and Water), razão gás óleo (RGO), unidade operacional de origem, entre outras. Os dados foram obtidos de um estudo retrospectivo de uma amostra com 450 poços-colunas que se encontravam em funcionamento no período de 2000 a 2006, escolhida de forma a representar todos os poços da bacia RN/CE. Foi realizada uma modelagem probabilística dos dados relativos à primeira falha através do ajuste do modelo de regressão Weibull. O modelo se mostrou apropriado para ajustar os dados e foi possível identificar, através do teste da razão de verossimilhança, quais e de que forma algumas características influenciam o tempo até a falha dos poços.

Quality Management and Failure Analysis Applied to Onshore Wells

“…About reliability and maintenance management are several tools that can be used for the purpose of this work. Dantas et al (2009) comments that "the reliability of equipment or industrial product is the likelihood that it successfully performs its functions, under a specific period of time, without a stop for maintenance or failure", being this concept important for Wells, because it is the sources of revenue from oil companies, already in relation to maintenance management we can highlight the use of quality management tools: flowchart, panser, Ishikawa diagram, check sheet, histograms, dispersion and letters Control as ways to enable the analysis, perdoná, naimer, souto and goday (2016).…”

“…In order to improve the life of the equipment used to raise fluids in the wells, it is necessary to apply failure analysis concepts, maintenance and reliability until a level of operating time compatible with the expected one reaches for each Method of elevation according to the literature of the sector, Dantas et al (2009). Sergipe and Alagoas Business Unit (Un-Seal) of Petrobras used these concepts and created a quality management system for failure analysis and increased time between wells from a computerized computer system with management tools maintenance and best practices of the international industry.…”

Time to Event Analysis for Failure Causes in Electrical Power Distribution Systems

“…For these types of studies, Reliability Analysis Regression (also know as Survival Analysis Regression) (Colosimo and Giolo, 2006) (Cox and Oakes, 2018) provides models to estimate the reliability and hazard using time as variable, making possible to determine the group of failure that is more recurrent, unlike an analysis considering only the number of failure events for each group. This approach is used in areas such as Medicine (Moolgavkar et al, 2018) and Engineering, in general problems (Dantas et al, 2010) or more specifically, as in Reliability Engineering (Murthy et al, 2004). For these models (Mishra et al, 2019), three major techniques can be found in the literature: the parametric (Zhang, 2016), the non-parametric (Rink et al, 2013),and the semi-parametric (Shauly et al, 2011).…”

Survival analysis of Electrical Power Distribution systems using Weibull Regression

“…Survival analysis grew in importance after World War II, in studies about the reliability of military equipment (Singh and Mukhopadhyay, 2011). The option of Survival analyses have three main methods, non-parametric, parametric and semi-parametric, and can be applied in areas such as Medicine (Moolgavkar et al, 2018), Engineering (Dantas et al, 2010) and Reliability (Murthy et al, 2004). Parametric Statistics needs a type of probability distribution and can be more accurate than non-parametric techniques (Zhang, 2016).…”