International Journal of Petroleum and Geoscience Engineering

Petrophysical evaluation of shaly gas sand reservoirs is one of the most difficult problems. These reservoirs usually produce from multiple layers with different permeability and complex formation, which is often enhanced by natural fracturing. In this study, we propose a new model to predict porosity and permeability using derived data from NMR. The developed Artificial Neural Network (ANN) model uses the NMR T2 pin values, density and resistivity logs to predict porosity, and permeability for two test wells. The NN trained model has displayed good correlation with core porosity and permeability values, and with the NMR derived porosity and permeability in the test wells. This work focuses on determination of porosity (φDMR) from combination of density porosity, NMR porosity and permeability from NMR logs using Bulk Gas Magnetic Resonance Permeability (KBGMR). Neural network (ANN) technique is used to predict formation porosity and permeability using NMR and conventional logging data. Predicted porosity and permeability have shown a good correlation about 0.912 with core porosity and about 0.891 for permeability in the studied shaly gas sand reservoir.

The gas compressibility factor is a key parameter in determining natural properties. The most common sources of gas compressibility factor (Z) values are experimental measurements, equation of state, and empirical correlations. There are more than twenty correlations available with two variables for calculating the Z-factor from fitting Standing-Katz chart values in EOS or through the fitting technique. The theory of corresponding states dictates that the Z-factor can be uniquely defined as a function of reduced pressure and temperature. Natural gases frequently contain material other than hydrocarbon components, such as nitrogen, carbon dioxide, and hydrogen sulfide. Hydrocarbon gases are classified as sweet or sour depending on the hydrogen sulfide content. Both sweet and sour gases may contain nitrogen, carbon dioxide, or both. The compositions of most natural gases are hydrocarbon of the same family (paraffin hydrocarbons), so the correlation of this type is possible but containing non-hydrocarbon on the gases, make the prediction difficult. 

A quick look at the well in Malay Basin data on the suitability of the AVO attributes and how the modeling result and derivatives of the product can lead to the false estimation of hydrocarbon. In most cases the amplitude change in the Malay basin is very small with regards to the rock quality, however, with the calibration to the well information, some products are feasible for application in the region. AVO method is the most common method practiced in the industry to delineate hydrocarbon prospecting based on pre-stack/partial seismic data, with no exception in the Malay basin field. There are a lot of AVO attributes developed by various researchers around the world, however, the sensitivity of each attribute in identifying the fluid types at the targeted reservoir is different.  This paper discusses the study of the sensitivity of several AVO attributes in differentiating the fluid types through AVO modeling and fluid replacement on the data set from a well located in the Malay Basin field.

Determination of water saturation in sandstone is vital question to determine initial oil or gas in place in reservoir rocks. Water saturation determination using electrical measurements is mainly on Archie’s formula. Consequently, accuracy of Archie’s formula parameters affects rigorously water saturation values. Determination of Archie’s parameters a, m and n is proceeded by three techniques conventional, CAPE and 3-D. This work introduces the hybrid system of parallel self-organizing neural network (PSONN) targeting an accepted value of Archie’s parameters and consequently reliable water saturation values. This work focuses on Archie’s parameters determination techniques; conventional technique, CAPE technique and 3-D technique and then calculation of water saturation using current. Using the same data, hybrid parallel self-organizing neural network (PSONN) algorithm is used to estimate Archie’s parameters and to predict water saturation. Results have shown that estimated Arche’s parameters m, an and n are highly accepted with statistical analysis indicating that PSONN model has lower statistical error and higher correlation coefficient. This study was conducted using high number of measurement points for 144 core plugs from sandstone reservoir. PSONN algorithm can provide reliable water saturation values and it can supplement or even replace the conventional techniques to determine Archie’s parameters and thereby calculation of water saturation profiles.

Accurate predictions of flow pattern, liquid holdup, and pressure drop are essential factors for oil and gas wells analysis and production optimization. In the literature, there are several empirical correlations and mechanistic models for predicting pressure loss during multiphase flow in wellbores. This study presents a comparative and performance analysis of three empirical correlations and three mechanistic models. Open source real field well datasets were utilized to estimate the pressure drop using MATLAB scripts created for each of the investigated correlations and models. The performance of the investigated empirical correlations and mechanistic models is evaluated using statistical error analysis, graphical error analysis, and relative error trend analysis. The empirical correlations demonstrated the best estimation of the pressure drop according to the investigation results because of the large-scale data used in establishing the correlations and the modifications made. The mechanistic models demonstrated the worst performance prediction according to the investigation results because of severe under-prediction of the pressure drop by the mechanistic slug flow and churn flow models. A solution procedure is proposed in this study that would eliminate the issue of non-convergent solutions. More study is needed to modify and improve the mechanistic slug flow and churn flow models.

Optimization of well locations and/or well injection/production controls by the trial-and-error method has proven to be computationally expensive and timeconsuming since numerous reservoir simulation studies need to be conducted to arrive at an optimum solution. A computationally inexpensive approach that combines response surface models and optimization algorithms in the optimization process is presented. A two-dimensional heterogeneous reservoir model with an injector and a producer was developed in this study with a reservoir simulator. Seven independent parameters namely bottom-hole pressure of the producer, gas injection rate, surfactant concentration, location of the producer and injector in i and j directions respectively were used. Using the minimum and maximum values of the independent parameters, Box-Behnken Design Method was used to generate fifty-six simulation runs, which were used as input in conducting reservoir simulations to arrive at an output. The input and output datasets were analyzed using experimental design software to generate a response surface model showing the relationship between cumulative oil produced and the seven independent parameters. The model was validated using statistical error analysis, the results of which show the accuracy and reliability of the model in navigating the design space. A comparison was made between cumulative oil produced obtained from the three optimization approaches. Results showed that a coupled well placement and well injection/production control optimization approach resulted in a higher value of cumulative oil produced. This work shows that considering a coupled well placement and well injection/production control optimization approach is preferable during field development planning and can be implemented using proxy models and optimization algorithms.