Predicting the Productivity Enhancement After Applying Acid Fracturing Treatments in Naturally Fractured Reservoirs Utilizing Artificial Neural Network

2021 ◽  
Author(s):  
Amjed Mohamed Hassan ◽  
Murtada Saleh Aljawad ◽  
Mohamed Ahmed Mahmoud
Keyword(s):  
Neural Network ◽  
Correlation Coefficient ◽  
Fractured Reservoirs ◽  
Naturally Fractured Reservoirs ◽  
Percentage Error ◽  
Reservoir Properties ◽  
Productivity Improvement ◽  
Acid Fracturing ◽  
New Correlation ◽  
Naturally Fractured

Abstract Acid fracturing treatments are conducted to increase the productivity of naturally fractured reservoirs. The treatment performance depends on several parameters such as reservoir properties and treatment conditions. Different approaches are available to estimate the efficacy of acid fracturing stimulations. However, a limited number of models were developed considering the presence of natural fractures (NFs) in the hydrocarbon reservoirs. This work aims to develop an efficient model to estimate the effectiveness of acid fracturing treatment in naturally fractured reservoirs utilizing an artificial neural network (ANN) technique. In this study, the improvement in hydrocarbon productivity due to applying acid fracturing treatment is estimated, and the interactions between the natural fractures and the induced ones are considered. More than 3000 scenarios of reservoir properties and treatment parameters were used to build and validate the ANN model. The developed model considers reservoir and treatment parameters such as formation permeability, injection rate, natural fracture spacing, and treatment volume. Furthermore, percentage error and correlation coefficient were determined to assess the model prediction performance. The proposed model shows very effective performance in predicting the performance of acid fracturing treatments. A percentage error of 6.3 % and a correlation coefficient of 0.94 were obtained for the testing datasets. Furthermore, a new correlation was developed based on the optimized AI model. The developed correlation provides an accurate and quick prediction for productivity improvement. Validation data were used to evaluate the reliability of the new equation, where a 6.8% average absolute error and 0.93 correlation coefficient were achieved, indicating the high reliability of the proposed correlation. The novelty of this work is developing a robust and reliable model for predicting the productivity improvement for acid fracturing treatment in naturally fractured reservoirs. The new correlation can be utilized in improving the treatment design for naturally fractured reservoirs by providing quick and reliable estimations.

After-Closure Analysis To Identify Naturally Fractured Reservoirs

2006 ◽  
Vol 9 (01) ◽  
pp. 50-60 ◽  
Author(s):  
Simon T. Chipperfield
Keyword(s):  
Low Cost ◽  
Fractured Reservoirs ◽  
Theory And Practice ◽  
Naturally Fractured Reservoirs ◽  
Reservoir Quality ◽  
Well Testing ◽  
Natural Fractures ◽  
Reservoir Properties ◽  
Reservoir Volume ◽  
Naturally Fractured

Summary After-closure analysis (ACA) in homogeneous-matrix reservoirs provides a method for extracting critical reservoir information from pre-frac injection tests. This paper extends the theory and practice of ACA to identify the presence of productive natural fractures. Natural fractures are important to identify before conducting a stimulation treatment because their presence may require designs that differ from conventional matrix treatments. Literature shows that naturally fractured reservoirs are very susceptible to formation damage and require stimulation treatments to account for this issue. The historical problem, however, has been to confidently characterize the reservoirs pre-frac in terms of both the reservoir quality and the deliverability mechanism (fractures vs. matrix) before committing to these design specifications. This paper presents the results of a simulator used to analyze the mini-frac after-closure period to identify the presence of natural fractures. The simulation results are distilled into a field implementation methodology for determining the extent of natural fracturing and the formation reservoir quality. This methodology is also applied to a field case study to verify the practicality of the technique. Unlike previous mini-frac-analysis methods, this approach identifies natural fractures that are material to production and allows the engineer to distinguish them from "fissures" that are open only during injection and are not a production mechanism. Introduction Motivation for Identifying Natural Fractures. Identifying the presence of natural fractures is important for a broad range of reasons. On a field scale, realizing the presence of natural fractures can impact reserves estimation, initial well rates, production declines, and planned well locations. With respect to well completions, fractured reservoirs may necessitate a special stimulation approach. Because fractured reservoirs tend to produce from a relatively small reservoir volume (i.e., the fractures), these formations can be highly susceptible to damage (Cippolla et al. 1988). The literature shows that the use of foamed treatments (Cippolla et al. 1988), 100 mesh, and low gel loadings can be used to stimulate these reservoirs effectively. The literature also shows the disastrous results that can arise when damage-prevention steps are not taken (Cippolla et al. 1988). As a result, there is a definite need to identify natural fractures before a stimulation treatment so that the appropriate design decisions can be made. In the past, conventional well testing, such as pressure-buildup tests, has been used for determining the reservoir description. However, these techniques often prove costly both in terms of additional equipment requirements and delays in well on-line dates. In addition, conventional well testing may not be successful in low-permeability reservoirs because these wells may not flow at measurable rates before stimulation. These cost and reservoir limitations have forced the engineer to seek other low-cost methods for determining reservoir properties. One such option for acquiring these data is the use of a mini-frac injection test conducted before a stimulation treatment. The mini-frac analysis techniques available to provide estimates of the formation capacity (kh) and indications of the presence of natural fractures include preclosure and post-closure methods.

scholarly journals Development of Acid Fracturing Model for Naturally Fractured Reservoirs

2019 ◽  
Vol 34 (04) ◽  
pp. 735-748
Author(s):  
Assiya Ugursal ◽  
Ding Zhu ◽  
Alfred Daniel Hill
Keyword(s):  
Fractured Reservoirs ◽  
Naturally Fractured Reservoirs ◽  
Acid Fracturing ◽  
Naturally Fractured

Well Testing of Fracture Corridors in Naturally - Fractured Reservoirs (NFR)

2021 ◽  
Author(s):  
Yingying Guo ◽  
Andrew Wojtanowicz
Keyword(s):  
Simulation Model ◽  
Fractured Reservoirs ◽  
Global Model ◽  
Naturally Fractured Reservoirs ◽  
Well Testing ◽  
Exclusion Zone ◽  
Reservoir Properties ◽  
Naturally Fractured ◽  
Corridor Network ◽  
Recognition Technique

Abstract Geological folding/faulting may create naturally fractured reservoirs containing a semi-parallel system of sparsely-spaced fracture corridors. The pressure behavior of wells completed either in highly conductive corridors (fracture wells) or in the exclusion zone (matrix wells) would be quite different. In this study, a unique simulation model has been built for corridor type naturally fractured reservoirs by combining a local model of fracture well or matrix well with adjacent fracture corridor(s) and a “homogenized” global model of the remaining corridor network. The global model generalizes the corridor network using the single-porosity and radial permeability approach, which is verified as being sufficiently accurate. Pattern recognition technique is used to analyze diagnostic plots of pressure drawdown generated by simulated flow tests with commercial software (CMG). This study aims to build a new simulation model for corridor-type NFRs and apply the well testing technique to differentiate corridor-type NFRs from conventional NFRs, detect the well’s location, and estimate reservoir properties. This study also employs cumulative logit statistics to assess the accuracy of the estimated well-to-corridor distance.

Sign in / Sign up

Export Citation Format

Share Document