Döngüsel buhar enjeksiyonu (DBE) inceleme modeli geliştirme.

Cyclic steam injection (CSI), a single-well enhanced oil recovery method for heavy oil reservoirs, is characterized with three stages: injection, soaking, and production which altogether constitute a cycle. In this study, it is aimed to develop a screening model that can be used to accept or reject a given CSI proposal from a representative performance indicator. This indicator is estimated from a large set of reservoir & CSI design characteristics, using the screening model developed. The model has been trained by using an artificial neural network (ANN) that can estimate the process performance in a given reservoir depending on the steam-injection design parameters. The data that be used for the ANN is generated using a representative numerical reservoir model, built with a commercial simulator. A large number of simulation cases are generated using the experimental design methodology to account for a large variety of scenarios, and corresponding performance indicators such as incremental oil recovery, and injection efficiency, are collected. After proper training and validation, the screening tool is ready to estimate the performance within a fraction of a second. Sensitivity study between the tool and numerical model showed that the tool captured the problem very well. According to the 90% of testing dataset results, the tool is able to estimate efficiencies with having less than 0.2 STB/STB absolute difference error. A probabilistic assessment study for a given reservoir illustrated the practicality of the tool.M.S. - Master of Scienc