No abstract

Selecting a proper enhanced oil recovery (EOR) method for a prospective reservoir is a key factor for successful application of EOR techniques. Reservoir engineers usually refer to screening guidelines to identify potential EOR processes for a given reservoir. However, these guidelines are often too general. In this study, we develop an advanced EOR screening technique based on the statistical analyses with boxplot in combination with some initial deep learning analyses to select the most suitable EOR method for a given mature oil field. At first, a database and the screening guidelines were established by compiling the information of 1,098 EOR projects from various publications in different languages, including Oil and Gas Journal (OGJ) biannual EOR surveys, SPE publications, DOE reports, and Chinese publications, etc. Boxplots were used to detect the special cases for each reservoir/fluid property and to present the graphical screening results. A case study was used to demonstrate that with a simple input of reservoir/fluid information, the proposed procedure could effectively give recommendations for EOR method selection. With the inputs (reservoir and fluid properties) from Vietnam offshore oil fields, the EOR methods recommended by this study are mostly chemical, including polymer and surfactant injection.

Applying the methods of enhanced oil recovery (EOR) for oil and gas fields has always many risks of economic and technology because EOR projects are influenced by many characteristic factors of the reservoir such as structure of reservoir, reservoir formation, geological properties, parameters of reservoir engineering, production technology to EOR application. Some EOR methods have been successfully applied in the world, but when these methods conduct in specific reservoir with different geological characteristics, tight production conditions have resulted in failures and ineffective economic, even caused dreadful aftermath to be handled in operations. Researches, evaluations and EOR applications in Vietnam are limited and only carried out on a laboratory scale. Therefore, the ability to be applied the EOR modern technology with a large scale or full field still faces many difficulties and the feasibility of projects is not high enough. The authors have been analysed all EOR projects successfully that applied many oil and gas fields in the world and then building EOR database. Based on EOR database, a study has been conducted on statistical analysis to build EOR screening criteria for reservoir parameters from past to now. The study also combined in-depth analysis algorithms such as Fuzzy, K - mean, PCA Artificial Intelligence to screen the optimal EOR method for sandstone reservoirs of Cuu Long Basin.