The demand for cost-effective drilling operations in oil and gas exploration is ever growing. One of the important aspects to tackling the aforementioned difficulty is determining the optimal rate of penetration (ROP) of the drill bit. The most important optimization objective is to achieve a high optimal rate of penetration in safe and stable drilling conditions. Several machine learning models have been developed to predict ROP, however, there have been few studies that consider the different optimization algorithms needed to optimize the conventional developed models other than the conventional grid search and random search techniques. Genetic algorithm (GA) has gained much attention as methods of optimizing the predictions of machine learning algorithms in different fields of study. In this study, GA optimization algorithm was implemented to optimize 5 machine learning algorithms: Linear Regression, Decision Tree, Support Vector Machine, Random Forest, and Multilayer Perceptron algorithm while using torque, weight on bit, surface RPM, mud flow, pump pressure, downhole temperature and pressure, etc, as input parameters. Three scenarios were analyzed using a train-test split ratio of 70-30, 80-20 and 85-15 percent on all the developed models. The results from the comparative study of all models developed shows that the implementation of the GA optimization algorithms increased the individual ROP models, with the multilayer perceptron model having the highest coefficient of determination of 0.989% after GA optimization.
Previous studies in literature and field experience indicates that hydrates mitigation can increase production cost by 15%. Hence, the need for optimum hydrates inhibitor selection. In this study, a deepwater oil field; currently producing at over 100,000 bbl/d had experienced hydrates and an integrated production modelling approach was utilized to model this scenario via MAXIMUS 6.20. The key input data to model were reservoir pressure at 2650 psia; Oil API Gravity at 35.4 °API; GOR at 1324 scf/stb and the produced water rate at an average of 3138 STB/d. Key results indicates that the injection of LDHI into the hydrate lattice within the hydrocarbon stream dispersed the hydrates particles in an unsteady manner and subsequently prevented the formation of hydrates at 40 gal/day. The cost of using LDHI was lower by over $520/day, from simulation results. This study also assessed technical benefits and challenges of using LDHI and MEG in deepwater scenario.
Oil based muds formulated from diesel oil have proven to be quite expensive and environmentally unfriendly. This current study focused on investigating the suitability of castor oil in formulating oil based muds. The experiment in this work was conducted at different temperature conditions ranging from 40 to 80°C. The density and specific gravity of castor oil based mud were found experimentally to be 0.3 ppg and 0.4 higher than the density and specific gravity of diesel oil based mud, respectively. Plastic viscosity and gel strength of castor oil based mud were found to be 2 cp and 2 lb/100 , respectively found to be less than those obtained for diesel oil based mud (15 cp and 4 lb/100 , respectively) indicating the ability of castor oil in enhancing rate of penetration and efficient in hole cleaning. Results also show that a thinner filter cake was obtained for castor oil based mud in comparison to diesel muds. This shows that with castor oil based mud, undesirable effects such as differential sticking, loss circulation and poor primary cement jobs are minimized. The result from toxicity test also implied that the drilling mud formulated with castor oil can be easily disposed with less post drilling treatment. In conclusion, oil based mud formulated with castor oil can be used as an alternative to diesel oil based mud for drilling operations because its properties show that it can effectively perform its functions as a drilling fluid. It is also environmentally friendly.
The Tertiary Niger Delta Basin is dominated by a friable, loosely consolidated sandstone formations indicating the likelihood of sand production ocuuring during hydrocarbon production in such formations. The aim of this study was to develop a simple mechanistic model for predicting sand production rate (SPR) in Niger-Delta wells. Two basic criteria (static sanding criteria and the dynamic requirement for fluidization of the produced sand) were used in developing this model. A generic mechanistic model that incorporates the concept of dimensionless quantities associated with sand prediction was developed. In developing the model, loading factor, Reynolds Number, water cut and gas-liquid ratio, GLR were considered. A dimensionless sand production rate (SPR) correlation index was the output from the proposed model. Results indicated that every reservoir has a unique SPR correlation index which represents its propensity to produce sand or its sanding identity. Validation of the proposed model was conducted by comparing the model prediction results with field data. Model validation results showed an agreement between predicted with field results with an acceptable maximum deviation of less than 5% in for onshore wells. The proposed model was further compared to existing models and prediction results show that the proposed model gave better results than others especially when the boost factor, GLR is significantly high. The applications of this study range from field development plans and economics to reservoir management, down to general well completion design and strategies.
Numerical reservoir simulation studies can be used to plan water injection projects to delay time and maximize oil recovery at water breakthrough which is time-consuming and computationally expensive. Combining computationally inexpensive proxy models and optimization algorithms is a solution to this problem. In this study, the Box-Behnken design method and response surface methodology were used to develop two proxy models which showed the relationship between time and recovery factor at water breakthrough with six independent variables namely porosity, horizontal permeability, water viscosity, bottom-hole pressure, water injection rate and vertical permeability. A comparison of actual and predicted values for time and oil recovery factor at water breakthrough was found to be in good agreement with each other. An average absolute percentage error of 2.038% and 1.217%, a root mean square error of 0.08 and 0.0000988, and coefficients of determination, R 2 of 0.9984 and 0.9946 were obtained for time and recovery factor at water breakthrough respectively. These are indications that the developed models are accurate, valid, and reliable. The models were further validated by comparing the actual and predicted water breakthrough time and recovery factor at water breakthrough using input variables that were not used in model development. These were also in close agreement with each other. The MATLAB multi-objective genetic algorithm was used to determine at a specific average porosity and permeability value, the best optimum controllable variables that maximized the objective functions. These were found to be 10.8978 years and 0.786 respectively and agreed with simulation results obtained using similar input parameter values.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.