The development of statistical analysis in the oil and gas industry database represents the importance of seeking improvements in safety and preventing undesirable events. For accident prevention, which may cause damage to the environment and financial losses, fault detection is essential. Modern techniques on the data-driven, such as data analysis and artificial intelligence are currently considered the best options for this purpose. Regardless of advances in techniques for database manipulation, scientists spend an amount of time working on data quality improvement. Thus, the high dimensionality of data introduces computational and statistical challenges such as acquisition, treatment, processing and interpretation of data. In this paper, we used the 3W public dataset published by Vargas et al. (2019), provided by Petróleo Brasileiro S.A. (Petrobras), which contains 8 variables such as pressure, temperature and flow rate in the process of offshore natural flow wells. The database was evaluated by data exploration and transformation that enables statistical analysis. The relationship between the variables was verified using histograms, Spearman's correlation and Principal Component Analysis (PCA). The results showed that at least one variable should be removed and others should be filled in order to complete the database. The analysis, also, revealed that the variables do not follow a normal distribution, and the variables importance rank. Thereby, it was possible to reach a database with useful format.