(2015036) AN EMPIRICAL MODEL TO PREDICT SKIN FACTOR OF A LIBYAN OILFIELD
Formation damage management and remediation are both a science and an art (Civan 1996). Currently, there are no proven technologies that are treated for all problems that an oil company may encounter. The issues revolving around formation damage is one of these convoluted issues which many oil companies currently struggle with. This paper has proposed such an innovative approach centered upon three dimensionless groups as well as multiple regression analysis using MINITAB (a statistical computing program) to foster an empirical model to predict skin factor for Field XXX which belongs to a Libyan Oil Company. The first step in this endeavor was employed by the use of data collection consisting of buildup data history and fluid properties from eight oil wells. A total of 39 observations were used in this study. Of these wells, 27 observations were used to develop the empirical model. The remaining 12 observations were chosen randomly to test the capability and validity of the model to validate the empirical model and test predictive competence, predicted skin factor values were compared against skin factor values determined from the buildup test analysis shown in Statistical evidence proved that the model illustrated in this thesis is advantageous and may potentially be utilized in efforts to predict of skin factor. Comparing the developed model predicted results to the observed buildup test results, demonstrations have shown that there is a correlation between the results and well ability of the developed model to estimate skin factor. As a result, this study offers the following conclusions: The size of the data set, used in the development of the empirical model, had significant effects on construction of the model, since the data used for developing the model must be good enough to increase the accuracy of model. In this study, 39 observations were used to form and test the model, which had six variables divided into three groups. These 39 observations represent five years of the production history of eight wells. The developed model presented in this study has the ability to further assist understanding, and evaluating the formation damage by predicting skin factor. The developed model also has the potential use of predicting skin factor instead of conducting a buildup test every year. This will reduce operating unit technical cost (UTC), and save millions of dollars for the Libyan operating company. When the mechanistic or mathematical models correlating certain variables are unknown, statistical tools are shown to be useful in development of models correlating with two or more variables of concern.