Multilateral well drilling technology has recently assisted the drilling industry in improving borehole contact area and reducing operation time, while maintaining a competitive cost. The most advanced multilateral well drilling method is Fishbone drilling (FbD). This method has been utilized in several hydrocarbon fields worldwide, resulting in high recovery enhancement and reduced carbon emissions from drilling. FbD involves drilling several branches from laterals and can be considered as an alternative method to hydraulic fracturing to increase the stimulated reservoir volume. However, the expected productivity of applying a Fishbone well from one field to another can vary due to various challenges such as Fishbone well design, reservoir lithology, and accessibility. Another challenge is the lack of existing analytical models and the effect of each Fishbone parameter on the cumulative production, as well as the interaction between them. In this paper, analytical and empirical productivity models were modified for FbD in a dry gas reservoir. The modified analytical model showed a higher accuracy with respect to the existing model. It was also compared with the modified empirical model, which proved its higher accuracy. Finally, machine learning algorithms were developed to predict FbD productivity, which showed close results with both analytical and empirical models.

Uniaxial Compressive Strength (UCS) and Tensile Strength (TS) are among the essential rock parameters required and determined for rock mechanical studies in Petroleum Engineering. However, the determination of such parameters requires some laboratory experiments, which may be time-consuming and costly at the same time. In order to estimate these parameters efficiently and in a short period, some mathematical tools have been used by different researchers. When regression tools proved to give good results only in the limited range of data used, machine learning methods proved to be very accurate in generating models that can cover a wide range of data. In this study, two machine learning models were used to predict the UCS and TS, Support Vector Regression optimized by Genetic Algorithm (GA-SVR) and Artificial Neural Networks (ANNs). The results were discussed for both uniaxial compressive strength and tensile strength in terms of coefficient of determination R2, root mean squared error (RMSE) and mean average error (MAE). First, for the case of UCS, values of 0.99 and 0.99, values of 3.41 and 2.9 and values of 2.43 and 1.9 were obtained for R2, RMSE and MAE for the ANN and GA-SVR, respectively. Second, for the TS, the same analogy was followed, a coefficient R2 of 0.99 and 0.99, RMSE values of 0.41 and 0.45 and MAE values of 0.30 and 0.39 were obtained for ANNs and GA-SVR, respectively. The next step was to assess these models on a different dataset consisting of data obtained from Bakken Field in Williston Basin, North Dakota, United States. The models showed excellent results comparing to the correlations they were compared with, outperforming them in terms of R2, RMSE and MAE, giving the following results for ANN and SVR respectively, R2 of 0.93, 0.92, RMSE of 9.54, 11.22 and MAE of 7.28, 9.24. The resultant conclusion of this work is that the use of machine learning algorithms can generate universal models which reduce the time and effort to estimate some complex parameters such as UCS and Tensile Strength.

One of the significant unconventional oil reserves in the USA is the Bakken Petroleum System located in the Williston Basin. It is known for its complex lithology, composed of three prominent members, Upper and Lower Bakken, with similar properties of organic-rich shale relatively uniform compared to the middle member with five distinct lithofacies, formed mainly from calcite, dolomite, or silica. The higher properties variability makes the reservoir characterization more challenging with low permeability and porosity. Understanding lithology by quantifying mineralogy is crucial for accurate geological modeling and reservoir simulation. Besides that, the reservoir’s capacity and the oil production are affected by the type and the mineral volume fractions, which impact the reservoir properties. Conventionally, to identify the mineralogy of the reservoir, the laboratory analysis (X-Ray Diffraction, XRD) using core samples combined with the well logs interpretation is widely used. The unavailability of the core data due to the high cost, as well as the discontinuities of the core section of the reservoir due to the coring failures and the destructive operations, are one of the challenges for an accurate mineralogy quantification. The XRD cores analysis is usually used to calibrate the petrophysical evaluation using well logs data because they are economically efficient. To remedy to these limitations, artificial intelligence and data-driven based models have been widely deployed in the oil and gas industry, particularly for petrophysical evaluation. This study aims to develop machine learning models to identify mineralogy by applying six different machine learning methods and using real field data from the upper, middle, and lower members of the Bakken Formation. Efficient pre-processing tools are applied before training the models to eliminate the XRD data outliers due to the formation complexity. The algorithms are based on well logs as inputs such as Gamma Ray, bulk density, neutron porosity, resistivity, and photoelectric factor for seven (07) wells. XRD mineral components for 117 samples are considered outputs (Clays, Dolomite, Calcite, Quartz, and other minerals). The results' validation is based on comparing the XRD Data prediction from the developed models and the petrophysical interpretation. The applied approach and the developed models have proved their effectiveness in predicting the XRD from the Bakken Petroleum system. The Random Forest Regressor delivered the best performance with a correlation coefficient of 78 percent. The rest of the algorithms had R-scores between 36 and 72 percent, with the linear regression having the lowest coefficient. The reason is the non-linearity between the inputs and outputs.

During the reservoir depletion and injection operations, the net effective stress is disrupted due to pore pressure changes. As a result, the reservoir properties, mainly porosity and permeability, are influenced by the change in the stress behavior in the reservoir rock. Understanding the porosity and permeability stress-dependent alteration is crucial since it directly impacts the reservoir storage capacity and the production/injection capabilities. Conventionally, lab experiments are conducted to understand the stress dependency of porosity and permeability magnitudes. Two methods are usually used: the unsteady-state method (Core Measurement System, CMS-300) and the steady-state method (Core Measurement System, CPMS). The challenges with these experiments reside in the fact that they are expensive and time-consuming and may cause the destruction of the core samples due to the applied stresses. This study aims to investigate the effect of stress variations on porosity and permeability changes. These properties were measured on a total of 2150 core data from the three members of the unconventional Bakken formation (upper, middle, and lower), applying 35 different Net Confining Stress (NCS) values, ranging from 400psi to 5800psi. A correlation was formulated between permeability and the NCS to illustrate the stress dependency relationships. The Grey Wolf Optimization algorithm (GWO) was used to tune the correlation for the Bakken formation. Machine Learning methods were also applied for the porosity and permeability stress dependency response prediction, which are as follows: Linear Regression (LR), Random Forest Regression (RF), XGBoost Regression (XGB), and Artificial Neural Network (ANN). The results demonstrate that the porosity and the permeability decrease with the increase of the NCS and vice versa. The permeability is highly sensitive to the NCS changes compared to the porosity. The developed correlations showed a good fit with the data extracted from the laboratory experiments of the pilot well. For the data-driven models, the coefficient of correlation R2-Score ranged from 91% to 93%. These models can be used to constrain the modeling work and reduce the uncertainties by introducing the effect of the net effective stress changes during reservoir depletion/injection on petrophysical properties.

Gas lift is one of the most commonly used artificial lift method in oil-producing wells. However, the technique requires constant optimization of gas allocation to maximize profit. The Gas Lift Performance Curves (GLPC) are the main design element that is used for optimized injection. Several authors have proposed models to fit the GLPC. These curves are generated by modeling wells in a multiphase steady-state simulator. Once the model is built, a sensitivity analysis is run, and the curves are generated. In this work, The common workflow to generate GLPC was followed. Then, a new correlation for GLPC was suggested. The correlation outperforms all the models in the literature in terms of R-score and root mean square error. The correlation was then used to formulate a case study for four wells located in North Africa. First, the wells and PVT models were used to create a simulation. Once the simulation was calibrated, a sensitivity analysis of the gas lift injection rate was run. The new correlation was used to fit the GLPC. The optimization problem was mathematically formulated, and stochastic optimization techniques were used, noting Grey Wolf Optimization (GWO) Algorithm and Genetic Algorithm (GA) to obtain the global optimum of the distribution of a limited gas lift quantity. Both algorithms’ results were compared. GWO slightly outperformed GA. The advantages of GWO over GA were discussed, and the optimum gas allocation was obtained.