In order to be able to accomplish steps 2 and 3 in the above process, the reservoir model must have the capability of analyzing multiple scenarios in real-time (or near real-time) and provide real-time responses to changes to the model input or potential modifications that can be made to the well operation. The reservoir/well responses to the modifications are reflected in high frequency (real-time) data streams. Above figure is a schematic diagram of the closed loop Real-Time Reservoir Management (RTRM™) concept.

Surrogate Reservoir Model: Surrogate Reservoir Model(Mohaghegh, 2009, 2006a, 2006b, 2006c)has been developed in response to the need for real-time reservoir modeling and in order to make Real-Time Reservoir Management (RTRM™) a reality. SRM is developed using the state-of-the-art in Artificial Intelligence & Data Mining (AI&DM). Artificial Intelligence & Data Mining is a collection of complementary analytical tools that attempt to mimic life when solving non-linear, complex and dynamic problems. AI&DM is consisted of, but is not limited to, analytical techniques such as Artificial Neural Networks, Genetic Optimization, and Fuzzy Logic.

Surrogate Reservoir Models (SRM) are accurate replicas of complex reservoir simulation models that may include tens or hundreds of wells. SRM runs provide results such as wells' pressure and production profiles or pressure and saturation distribution throughout the reservoir, in real time. SRM is developed using a unique and proprietary series of data generation, manipulation, compilation and management techniques. These techniques are designed to take the maximum advantage of characteristics of artificial neural networks complemented with fuzzy set theory. Upon completion of modeling process and validation, SRM can accurately replicate the results generated by highly sophisticated reservoir simulation models in respond to changes made to the model input, in fractions of a second. The fact that SRM runs in real-time makes (near real time) uncertainty analysis possible so the uncertainty band associated with the decisions that are made can be identified.

SRM has been field tested. In a recent study performed on a giant oil field in the Middle East, a SRM was developed to replicate the existing simulation model of the field that was developed using a commercial simulator. The computing time required for a single run of the existing simulation model is 10 hours on a cluster of 12 parallel CPUs. Upon development of the SRM that could successfully and accurately replicate the results of the simulation model, tens of millions of SRM runs were made in order to comprehensively explore the solution space of the reservoir model in order to develop a field development strategy. The objective was to increase oil production from the field by relaxing the rate limitation on wells. The key was to identify those wells that will not suffer from high water cut once the rate relaxation program is put into place. The SRM had to take into account and quantify the uncertainties associated with the geological model while accomplishing the objectives of this project.

Upon completion of tens of millions of SRM runs (equivalent to tens of millions of simulation runs) the 165 wells in the field were divided into 5 clusters. It was recommended that wells in clusters 1 and 2 be subjected to rate relaxation. Furthermore, it was predicted that these wells will produce small amount of water and large amount of incremental oil in the next 25 years. On the other hand, more than 100 wells that were placed in clusters 4 and 5 were identified as wells that will produce large amounts of water in case the rate restrictions were to be lifted.

Upon completion of the study, rate restriction was lifted from 20 wells. These wells were selected from among all the clusters to provide a representative spatial distribution of the reservoir. After more than two and a half years of production the results were analyzed. As can clearly be seen in the above figures, wells in clusters 1 and 2 produced large amounts of incremental oil while the water production reduced. The opposite effect was observed in wells that were classified in clusters 4 and 5, as predicted by SRM. Figure below shows the maximum incremental water cut normalized for all wells in each of the clusters. It is clear from this figure that in accordance with SRM's predictions water cut decreased in wells classified in clusters 1 and 2 while increase significantly in wells classified in clusters 4 and 5.

Results shown in the above study as well as other similar studies demonstrate the robustness of SRM technology. SRM can be used to develop replicas of sophisticated and large reservoir simulation models that can then be used in order to drive the main engine of Real Time Reservoir Management (RTRM™).

INTELLIGENT REAL-TIME DATA ANALYSIS : The high frequency (real-time) data that is collected from the permanent downhole gauges and transmitted to be stored in data historians is usually unusable in its raw form. It needs to be cleansed and summarized and prepared (processed) for use in reservoir engineering studies. The high frequency data needs to be de-noised, the outliers must be identified and removed, existing trends and patterns need to be identified and data need to be summarized so that maximum information can be preserved using the least amount of data. Most importantly all these need to be performed reliably, in real-time (at the same time scale - or faster - that data is received) automatically and autonomously without supervision of an engineer.

Furthermore, the intelligent real-time data analyzer needs to have capabilities of taking maximum advantage of the information content of the high frequency data. ISI's intelligent data analysis of the high frequency data streams include:

1. High Frequency Data Management.
   High frequency data management performs all preparations and preprocessing of the data autonomously and in real time. During this process data is prepared to be used in reservoir engineering analysis and it includes:
   • Data de-noising
   • Outlier removal
   • Pattern recognition and summarization
   • Data preparation for SRM

2. High Frequency Data Analysis.
   High frequency data analysis performs state-of-the-art reservoir engineering analysis using the high frequency data streams:
   • Detection, isolation and analysis of pressure transient data used for continuous monitoring of well and reservoir behavior,
   • Modeling using adaptive technology that learns data behavior and continuously modifies itself to match and model observed data and to predict its behavior,
   • Detect, model and verify hypotheses about drive mechanisms by continuous modeling of volumetric estimates
   • Performing diagnostic analysis in order to detect inter-well connectivity that may or may not exist between multiple wells in a reservoir, and finally
   • Communication (send feedback, receive instructions) with the SRM.

REFERENCES
Intelligent Data Evaluation & Analysis; IDEA™
Project: Using SRM for Fast Track Analysis of a Giant Oil Field in the Middle East
SPE 99667
SPE 101474
SPE 102492