document preview

Abstract
Production data analysis has extensively been applied to predicting the future production performance and field recovery. These applications are mostly on a single well basis. This paper presents a new approach to production data analysis using Artificial Intelligence (AI) techniques where production history is used to build a field-wide performance prediction models. In this work AI techniques and data driven modeling are utilized to predict future production of both synthetic (for validation purposes) and real field cases. 

In the approach presented in this article production history is paired with geological information from the field to build datasets containing the spatio-temporal dependencies amongst different wells. These dependencies are addressed by compiling information from Closest Offset Wells (COWs) that includes their geological and reservoir characteristics (spatial data) as well as their production history (temporal data). 

Once the dataset is assembled, a series of neural networks are trained using backpropagation algorithm. These networks are then fused together to form the “Intelligent Time-Successive Production Modeling“(ITSPM) system. This technique only uses the widely available measured data such as well logs and production history of existing wells to predict a) future performance of the existing wells, b) production performance of the new (infill) wells and c) the initial hydrocarbon in place using a “volumetric-geostatistical” method. 

To demonstrate the applicability of this method, a synthetic oil reservoir is modeled using a commercial simulator. Production and well-log data are extracted into an all-inclusive dataset. Several neural networks are trained and validated to predict different stages of the production. ITSPM method is utilized to estimate the production profile for nine new wells in the reservoir. Furthermore, ITSPM is also applied to a giant oil field in the Middle East with more than 200 wells and forty years of production history. ITSPM’s production predictions of the four newest wells in this reservoir are compared their actual production.

Introduction 
Two of the most influential pieces of information in decision making and field developments are our depth of knowledge about the reservoir’s state of depletion and the estimation of the remaining reserves. These become more important in brown fields which most of wells are in their decline period and new wells can easily become non profitable if not drilled in the right spots. 

There are several techniques enabling the reservoir engineers to have a reservoir model that is capable of predicting future behavior of the reservoir under different development strategies. These models are normally based on numerical solutions of the fluid flow equation and require fairly accurate data and measurements about the formation. Furthermore, acquiring such data and measurements are expensive and building full field models to use these data in a cohesive manner can be financially prohibitive considering the computational and human resources required. 

In contrast, instead of lengthy and expensive numerical solutions, analytical solutions are simpler and cheaper. These solutions are normally limited to single well based analyses with many limiting homogeneity assumptions. Although these solutions are much easier to develop and they do not need vast amount of data nor computer power, their deliverability is limited. 

Relying on availability of large amount of data and measurements about the field is not always a realistic solution. Therefore the numerical solutions are not always practical. Moreover, single well analysis techniques are not always good choices for field development strategy and decision makings. 

Brown fields with marginal production rates or old fields without state-of-the-technology studies, are not the best candidates for costly numerical simulation models. In some cases single well numerical models are built for some fields; these models limit the analysis to one well and don’t give a full field understanding of the reservoir.