Prediction of uniaxial compressive strength of rock samples using hybrid particle swarm optimization-based artificial neural networks

Many studies have shown that artificial neural networks (ANNs) are useful for predicting the unconfined compressive strength (UCS) of rocks. However, ANNs do have some deficiencies: they can get trapped in local minima and they have a slow learning rate. It is widely accepted that optimization algorithms such as particle swarm optimization (PSO) can improve ANN performance. This study investigated the application of a hybrid PSO-based ANN model to the prediction of rock UCS. To prepare a dataset for the predictive model, extensive laboratory tests (i.e., 160 tests in total) were conducted on 40 soft rock sample sets (mostly shale) presenting various weathering grades that were obtained from different excavation sites in Johor, Malaysia. The laboratory tests included the UCS test and other basic rock index tests (the Brazilian tensile strength test, point load index test, and ultrasonic test). When developing the predictive model of UCS, the results of the basic rock tests as well as the bulk densities of the samples were used as input parameters, while the UCS was set as the output of the predictive model. The value account for (VAF), root mean squared error (RMSE), and adjusted R
 2 (coefficient of determination) were utilized to check the performances of the predictive models. The high performance indices of the proposed model highlight the superiority of the PSO-based ANN model for UCS prediction.

Prediction of the unconfined compressive strength of soft rocks: a PSO-based ANN approach

Use of machine learning and data analytics to increase drilling efficiency for nearby wells

Real-time predictive capabilities of analytical and machine learning rate of penetration (ROP) models

Artificial intelligence techniques and their applications in drilling fluid engineering: A review

Intelligent Prediction of Differential Pipe Sticking by Support Vector Machine Compared With Conventional Artificial Neural Networks: An Example of Iranian Offshore Oil Fields

Real-time Prediction of Rate of Penetration During Drilling Operation In Oil And Gas Wells

Accurate prediction of pore pressures has become almost essential to exploration, completion and drilling wells with higher-than-normal pore pressures. There are some limitations with the current existing pore pressure estimation methods, as they are based on empirical relations and constants which can differ from basin to basin. In this study, a feed-forward network with back-propagation algorithm and a generalised regression neural network have been developed to predict pore pressure gradient in Asmari reservoir in one of oilfields in Iran. To show the accuracy of the method, the results of the developed models have been compared with Eaton’s method.

Real-time prediction of pore pressure gradient through an artificial intelligence approach: a case study from one of middle east oil fields

Lost circulation is one of the most troublesome and costly problems in drilling industry. This phenomenon can lead to some serious problems such as wellbore instability, differential pipe sticking and blow out. Meanwhile, it is possible to mitigate the risk of lost circulation by carrying out geomechanical analysis and real-time prediction. Artificial neural network (ANN) is an efficient tool to predict the events due to the effective parameters. Generally, there are various factors including geomechanical and non-geomechanical parameters that affect lost circulation. In this paper, two ANNs have been developed to predict the lost circulation in naturally fractured formations. Firstly, an ANN has been designed without considering geomechanical parameters as inputs. Then another ANN has been developed by considering both the geomechanical and non-geomechanical parameters. Finally, two ANNs have been compared to investigate the geomechanical parameters effect on lost circulation. In both the models, princip...

Artificial neural network-based prediction and geomechanical analysis of lost circulation in naturally fractured reservoirs: a case study

Hydraulic fracturing of naturally fractured reservoirs is a critical issue for petroleum indus‐ try, as fractures can have complex growth patterns when propagating in systems of natural fractures. Hydraulic and natural fracture interaction may lead to significant diversion of hy‐ draulic fracture paths due to intersection with natural fractures which causes difficulties in proppant transport and eventually job failure. In this study, a comparison has been made between numerical modeling and artificial intelligence to investigate hydraulic and natural frcature interaction. First of all an eXtended Finite Element Method (XFEM) model has been developed to account for hydraulic fracture propagation and interaction with natural frac‐ ture in naturally fractured reservoirs including fractures intersection criteria into the model. It is assumed that fractures are propagating in an elastic medium under plane strain and quasi-static conditions. Comparison of the numerical and experimental studies results has shown good agreement. Secondly, a feed-forward with back-propagation artificial neural network approach has been developed to predict hydraulic fracture path (crossing/turning into natural fracture) due to interaction with natural fracture based on experimental studies. Effective parameters in hydraulic and natural fracture interaction such as in situ horizontal differential stress, angle of approach, interfacial coefficient of friction, young’s modulus of the rock and flow rate of fracturing fluid are the inputs and hydraulic fracturing path(cross‐ ing/turning into natural fracture) is the output of the developed artificial neural network. The results have shown high potentiality of the developed artificial neural network ap‐ proach to predict hydraulic fracturing path due to interaction with natural fracture. Finally, both of the approaches have been examined by a set of experimental study data and the re‐ sults have been compared. It is clearly observed that both of them yield promising results © 2013 Keshavarzi and Jahanbakhshi; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. while numerical modeling yields more detailed results which can be used for further inves‐ tigations but it is computationally more expensive and time-consuming than artificial neural network approach. On the other hand, since artificial neural network approach is mainly da‐ ta-driven if just the input data is available (even while fracturing) the hydraulic fracture path (crossing/turning into natural fracture) can be predicted real-time and at the same time that fracturing is happening.

/pdf/investigation-of-hydraulic-and-natural-fracture-interaction-4xhfgztvi3.pdf

R. Jahanbakhshi

Papers

Intelligent Prediction of Differential Pipe Sticking by Support Vector Machine Compared With Conventional Artificial Neural Networks: An Example of Iranian Offshore Oil Fields

Real-time Prediction of Rate of Penetration During Drilling Operation In Oil And Gas Wells

Real-time prediction of pore pressure gradient through an artificial intelligence approach: a case study from one of middle east oil fields

Artificial neural network-based prediction and geomechanical analysis of lost circulation in naturally fractured reservoirs: a case study

Investigation of Hydraulic and Natural Fracture Interaction: Numerical Modeling or Artificial Intelligence?