International Journal of Communication and Information Technology

P-ISSN: 2707-661X, E-ISSN: 2707-6628
Printed Journal   |   Refereed Journal   |   Peer Reviewed Journal

2021, Vol. 2, Issue 1, Part A

Prediction of properties of reservoir using Artificial Neural Network and Monte Carlo Simulation


Author(s): Adeboye Olatunbosun, OA Falode and Engr Awani Kester

Abstract: Two methods for forecasting reservoir distribution are used: Monte Carlo Simulation and Artificial Neural Network. The data used in this research was obtained from the drilling sheet report of the Efe04 well located in the Northern Niger Delta Depobelt. The operator released the data set of the Efe field. Amukpe well data from the Amukpe field was also used as part of the test data to see how the model works on different fields. To minimize uncertainties, only the surface drilling parameters are used for the model. Mode of operation (rotation or sliding), Torque, Surface Revolution, Flow rate (gallons per minute), Stand Pipe Pressure, Revolutions per Minute, hole size. When reservoir distribution can be predicted, it is extremely beneficial since it increases exploration accuracy and lowers costs by delaying the next round of exploration. Geostatistics, on the other hand, is rarely used in areas with few wells drilled. A new technique called Geology Driven Integration Tool (GDI) has been developed to estimate reservoir parameters when just a few wells are available. Due to the lack of well data and regional geological constraints in the GDI model, several pseudo-wells are constructed by Monte Carlo Simulation to make up for the deficiency of a few real wells. They can also be used to create fake seismograms. To determine the weighting factors that link the selected seismic attributes to the given reservoir features, the appropriate seismic attributes and the given reservoir parameters are input to the Artificial Neural Network (ANN). In the end, the ANN is trained on all of the seismic data in the area and then used to estimate reservoir property distribution. In addition to the areas that have already been explored, the estimated results suggest exploring the southern portion of the Efe field and the northern portion of the Efe field as potential prospect locations. In the southern half of the Amukpe field, the gas zone's net thickness is expected to rise to 27 meters, thanks to increased porosity of 27%. North of the Efe field, a 15-25-meter-thick porosity-rich reservoir is predicted to be distributed.

DOI: 10.33545/2707661X.2021.v2.i1a.32

Pages: 36-41 | Views: 644 | Downloads: 363

Download Full Article: Click Here
How to cite this article:
Adeboye Olatunbosun, OA Falode, Engr Awani Kester. Prediction of properties of reservoir using Artificial Neural Network and Monte Carlo Simulation. Int J Commun Inf Technol 2021;2(1):36-41. DOI: 10.33545/2707661X.2021.v2.i1a.32
International Journal of Communication and Information Technology

International Journal of Communication and Information Technology

International Journal of Communication and Information Technology
Call for book chapter