|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
ABSTRACT: Asmari and Sarvak limestones are two main oil producer formations in Iran and the Middle East. The production and optimal utilization of these reservoirs will have a significant impact on the economy of the petroleum industry. Geomechanical modelling of oil reservoirs are widely used in optimum drilling, production and reservoir compaction. Hence, the static Young’s modulus (Es) is one of the most essential parameters for any reservoir geomechanical modelling. However, information on the values of Es along the well depth is often discontinuous and limited to the core locations. Therefore, dynamic Young’s modulus (Ed) determined from open-hole log data such as density and compressional and shear wave velocities could result in continuous estimation of elastic properties of the formations versus depth. Nevertheless, static parameters are more reliable than the dynamic parameters and they are widely accepted by geomechanics around the world. The relationship between the static and dynamic elastic modulus in rock materials has been frequently addressed in scientific literature. Overall, when it comes to the study of materials with a wide range of elastic moduli, the functions that best represent this relationship are non-linear and do not depend on a single parameter. Therefore, finding a valid correlation between static and dynamic parameters could result in a continuous and more reliable knowledge on elastic parameters. In this study, published data of the tests which were carried out on 45 Asmari and Sarvak limestone core specimens are used. Then, as an artificial intelligence method, artificial neural networks were developed to correlate Es and Ed data. After comparing the results of the suggested method with correlations which were established between dynamic and static measurements, a good agreement was observed. The accuracy of the obtained results have shown that artificial neural networks are appropriate tools to predict the values of Es based on Ed data of limestone formations.
Abstract It is known that pore pressure (Pp) is an integral part for the well planning process. Pore pressure can be directly measured from the wireline pressure and well tests, or indirectly measured from seismic velocities, well logs, and shale densities. While the direct measurements are limited due to the cost and time-saving purposes, the indirect methods are often used, especially the techniques that based on the mechanical compaction of fine-grained sediments. However, the loss of porosity in carbonate reservoirs is not only controlled by the effective stresses, but also affected by a variety of depositional environments and diagenetic processes. Most of the previous models were developed to detect the overpressure zones rather than the subnormal (i.e., depleted) zones. There are also some limitations in the traditional methods, as they are based on empirical relations and constants that can differ from basins to others. This study presents a regression analysis (RA) and artificial neural networks (ANNs) capable of predicting the Pp using measurable well logs. A field case, located in SE Iraq, has been investigated to determine the Pp from well log data. A database for five offset wells of Mishrif reservoir was subjected to the predictive methods. Two traditional methods, the Eaton and the Ratio methods, were also conducted to compare their performance with in-situ pore pressure data in carbonate reservoirs. The current results showed that the true vertical depth, bulk density, neutron porosity, gamma ray, compressional travel time, and unconfined compressive strength are the key parameters for the Pp prediction. An empirical model with a good performance using ANNs has been developed to estimate the Pp using petrophysical well logs. Although both RA and ANNs are conservative in predicting Pp, the higher value of determination coefficient (0.96) of ANNs demonstrated that the ANN can predict the subnormal pore pressures in carbonate reservoirs. While the Eaton and the Ratio methods which are based on the drilling derived dc values showed a closer alignment with the in-situ Pp direct measurements, they are not applicable in depleted carbonate reservoirs. Other indicators of the prediction Pp should be used in conjunction with penetration rate. The validity of the proposed models was successfully checked with the data from another field study in SE Iraq. This study presents efficient and cost-effective models to estimate the formation pore pressure in depleted carbonate environments utilizing petrophysical well logs.
Summary Estimates of formation pore pressure before and while drilling are important inputs for well planning and operational decision making. A method is proposed to determine pore pressure from a combination of downhole drilling-mechanics parameters and in-situ rock data with the concept of mechanical specific energy (MSE) and drilling efficiency (DE). This pore-pressure estimation method (termed DEMSE) is based on the theory that energy spent at the bit to remove a volume of rock is a function of in-situ rock strength and the differential pressure that the rock is subjected to during drilling. A work flow is provided that illustrates the steps required to estimate pore pressure from drilling parameters and rock-mechanics data by use of the DEMSE method. Pore pressure estimated from the DEMSE method is compared with pore-pressure estimates derived through a conventional sonic log that is based on empirical technique for a deepwater well in the Gulf of Mexico (GOM). Pore-pressure estimates from the DEMSE method generally agree in magnitude and trend with the pore-pressure estimates derived from sonic-log data. The results of the DEMSE method have also been compared with pore-pressure estimates from the classical d-exponent (dXc) approach to highlight the advantages of DEMSE over traditional dXc methods. Finally, the importance of using downhole vs. surface data for pore-pressure estimation purposes, specifically torque measurements at the bit, is illustrated through a field example. These findings suggest that downhole drilling-mechanics data, when properly used, can provide reliable independent estimates of pore pressure in real time at the bit and can be used for post-well-analysis to assist with constructing pore-pressure forecasts.
Abstract Recent studies observe drilling optimization in real-time, however, among the investigated references, there is no practice working with pump pressure prediction by considering formation variables such as depth and independent parameters like Revolutions Per Minute (RPM), Hook Load. Reliable prediction of pump pressure provides an early warning of circulation problems, washout, underground blowout, and kicks helping the driller to make corrections and to safely avoid major problems. Throughout this particular study, an Artificial Neural Network model was implemented through the fitting tool of MATLAB. This model can accurately predict pump pressure versus depth in similar formations. Following the determination of the optimum model the sensitivity analysis of input parameters on the created model was investigated, an overall ranking of sensitivity degree was then provided to show the impact of each individual input parameter on this model. The simulation result was promising. Therefore, the result of this work shows the potential of a neural network approach to model the hydraulic behavior in a well. Hopefully, this work can be included in a software system which can be used in drilling planning and real-time operation of oil and gas wells in the related field that can result in decreasing Non Productive Time.
Amadi, Kingsley Williams (Australian College of Kuwait) | Iyalla, Ibiye (Robert Gordon University Aberdeen) | Liu, Yang (University of Exeter, England) | Alsaba, Mortadha (Australian College of Kuwait) | Kuten, Durdica (Australian College of Kuwait)
Abstract Fossil fuel energy dominate the world energy mix and plays a fundamental role in our economy and lifestyle. Drilling of wellbore is the only proven method to extract the hydrocarbon reserves, an operation which is both highly hazardous and capital intensive. To optimize the drilling operations, developing a high fidelity autonomous downhole drilling system that is self-optimizing using real-time drilling parameters and able to precisely predict the optimal rate of penetration is essential. Optimizing the input parameters; surface weight on bit (WOB), and rotary speed (RPM) which in turns improves drilling performance and reduces well delivery cost is not trivial due to the complexity of the non-linear bit-rock interactions and changing formation characteristics. However, application of derived variables shows potential to predict rate of penetration and determine the most influential parameters in a drilling process. In this study the use of derived controllable variables calculated from the drilling inputs parameters were evaluated for potential applicability in predicting penetration rate in autonomous downhole drilling system using the artificial neutral network and compared with predictions of actual input drilling parameters; (WOB, RPM). First, a detailed analysis of actual rock drilling data was performed and applied in understanding the relationship between these derived variables and penetration rate enabling the identification of patterns which predicts the occurrence of phenomena that affects the drilling process. Second, the physical law of conservation of energy using drilling mechanical specific energy (DMSE) defined as energy required to remove a unit volume of rock was applied to measure the efficiency of input energy in the drilling system, in combination with penetration rate per unit revolution and penetration rate per unit weight applied (feed thrust) are used to effective predict optimum penetration rate, enabling an adaptive strategize which optimize drilling rate whilst suppressing stick-slip. The derived controllable variable included mechanical specific energy, depth of cut and feed thrust are calculated from the real- time drilling parameters. Artificial Neutral Networks (ANNs) was used to predict ROP using both input drilling parameters (WOB, RPM) and derived controllable variables (MSE, FET) using same network functionality and model results compared. Results showed that derived controllable variable gave higher prediction accuracy when compared with the model performance assessment criteria commonly used in engineering analysis including the correlation coefficient (R2) and root mean square error (RMSE). The key contribution of this study when compared to the previous researches is that it introduced the concept of derived controllable variables with established relationship with both ROP and stick-slip which has an advantage of optimizing the drilling parameters by predicting optimal penetration rate at reduced stick-slip which is essential in achieving an autonomous drilling system. :