The use of artificial neural networks to recognize the model of horizontal wells in oil reservoirs using well test data

Master thesis in chemical engineering (gas engineering orientation)

In recent years, many horizontal wells have been drilled around the world. The main reason is the ability to increase the level of the reservoir in contact with the well, which increases the utilization of the well. Well testing is used to identify the models of hydrocarbon reservoirs and to identify their related parameters. Well testing works on the basis of disrupting the flow and recording the pressure in the bottom of the well caused by it. This technique provides the data required for the numerical analysis of reservoir parameters. The well testing method includes two steps: 1) Reservoir model classification 2) Parameter estimation. Identifying horizontal well models and determining the parameters of their models is much more complicated compared to vertical wells. Determining the reservoir model from pressure derivative diagrams is one of the important and basic steps in estimating reservoir parameters through the analysis of well test data. In this study, artificial neural networks have been used to identify the model of oil reservoirs through pressure derivative diagrams. Artificial neural networks are mathematical models that have a unique ability to estimate parameters and identify patterns. are Eight different horizontal well models of oil reservoirs, which include homogeneous and dual porosity reservoirs with different boundaries, have been investigated. The forward neural network is trained by pressure derivative data generated by simulation with PANSYSTEM software. The performance of the perceptron network is checked by means of the average relative errors and the square of the mean squared error. The ability of the designed network has been investigated through noisy data. The accuracy of the network is given by a number of statistical parameters such as the sensitivity and accuracy of the overall classification, and the overall accuracy of the leading network is 05/97. rtl;">Introductions to Reservoir Engineering

Crude oil, natural gas and water are materials that are of particular importance to petroleum engineers. These substances, which are sometimes found in solid or semi-solid form at low temperature and pressure (such as paraffin, gas hydrates, ice, and crude oil with a high pouring point), appear deep in the ground and in the well column in a liquid state, in the form of a vapor (gas) or liquid phase, or mainly in two phases. Solid materials that are used in drilling, cementing and creating gaps are also used in liquid or slurry form. The division of reservoir and well fluids into liquid and vapor phases depends on temperature and pressure. When the temperature is constant, the state or phase of the fluid inside the tank changes with the pressure. In many cases, the state or phase of the fluid in the reservoir does not match the state or phase of the fluid during production at surface conditions. Accurately understanding the behavior of crude oil, natural gas and water - singly or in combination - under different conditions is one of the most important goals of petroleum engineers. 

Early in 1928, special attention was paid to gas and energy relations, and petroleum engineers found it necessary to obtain more accurate information about the physical conditions of wells and underground reservoirs. Early advances in oil recovery methods revealed that calculations based on wellhead or surface data are often misleading. Sclater and Stephenson[1] invented the first downhole pressure recorder and sampler for sampling fluids under pressure in wells[1]. Interestingly, this device determines the data inside the well according to the positive values ??of pressure, temperature, gas-oil ratios and the physical and chemical nature of fluids. The need to measure the correct pressures inside the well was noticed when the first accurate pressure gauge was made by Millikan and Seidol [2] and the essential importance of the pressures inside the well in determining the most effective recovery methods and extraction processes was shown to petroleum engineers [2]. In this way, the reservoir engineer will be able to measure the reservoir pressure, which is the most important basic data required for reservoir performance calculations.

Petrophysics knowledge is the study of the properties of rocks and the relationship with the fluids in them in both static and flowing states.Porosity, permeability, degree of saturation and distribution of fluids, electrical conductivity coefficient of rock and fluid, pore structure and radioactivity are some of the most important petrophysical properties. The pioneers of reservoir engineering science realized from the very beginning that before calculating the volumes of oil and gas in situ, it is necessary to know the change of the physical properties of the bottom well samples of the reservoir fluids, relative to the pressure. These terms are synonymous and refer to the ability to use mathematical equations to predict the performance of an oil or gas reservoir. The emergence of high-speed digital computers on a large scale strengthened the science of reservoir simulation. Complex numerical methods were also developed using finite difference or finite element methods to solve a large number of equations.

With the development of these methods, concepts and equations of reservoir engineering became strong defined branches of petroleum engineering. Reservoir engineering is the application of scientific principles to solve drainage problems that arise during the development and operation of oil and gas reservoirs. Reservoir engineering (the art of developing and exploiting oil and gas fluids in a way that achieves high economic recovery) has also been defined [4]. Fortunately, oil and gas masses are usually found in the more porous and permeable parts of the beds, which are mainly sands, sandstones, limestones, and dolomites, and also in the interests between grains or pore spaces created by seams, cracks, and solution activity.    

In the initial conditions of the reservoir, hydrocarbon fluids are in single phase or two phase state. The single phase state may be the liquid phase where all the gas in the oil is dissolved. In this case, soluble natural gas reserves should be estimated in the same way as crude oil reserves. Alternatively, the single-phase state may be the gas phase. If there are vaporized hydrocarbons in the gas phase that can be recovered in the form of natural gas liquids on the surface of the earth, this reservoir is called a condensate gas reservoir or a distillate gas reservoir. In this case, the available accompanying liquid reserves (condensate or distillate) should be estimated in the same way as gas reserves. When a hydrocarbon mass is two-phase, the vapor phase is called a gas cap, and the liquid phase located below it is called an oil zone. There will be four types of hydrocarbon reserves here:

Free gas or associated gas, dissolved gas, oil in an oil zone and natural gas liquids that are recovered from the gas cap.

Although the hydrocarbons in the reservoir, which is called reserves, have fixed amounts, the amount of reserves depends on the method of exploitation of the reservoir. In 1986, the Society of Petroleum Engineers (SPE)[3] chose the following definition for reserves:

Reserves are the estimated volumes of crude oil, natural gas, natural gas liquids and associated materials that can be supplied in the market from a certain point of time under the existing economic conditions, with specific exploitation operations and under the current regulations of the government, from an economic point of view, have the ability to recover and profit and supply in the market [6]. The amount of reserves is calculated using available geological and engineering data. Gradually, as more data is obtained during the exploitation of the reservoir, the estimation of the reserves is also updated. In initial exploitation, oil or gas is pushed towards production wells due to a) expansion, b) fluid displacement, c) gravity fall and d) repulsive capillary force. If the tank does not have a water table and no fluid is injected into it, hydrocarbon fluid recovery is mainly done by fluid expansion. While in the case of oil, recovery may be done with the help of gravity fall mechanism.