Feasibility Study on the Applicability of Intelligent Well Completion

Abstract

The relevance of assessing the applicability of intelligent wells using autonomous inflow control devices lies in the active development of the relevant sector of the oil and gas industry and the limited understanding of the economic efficiency of intelligent wells. The use of autonomous inflow control devices allows for a change in the composition of flow to the well, thus contributing to delaying the breakthrough of undesirable formation fluids, but at the same time, such an effect affects the dynamics of formation fluid production, which undoubtedly has a huge impact on the economic effect of the project. The analysis of scientific publications on the topic of “intelligent well completion” as a downhole production monitoring and remote production control system has shown that the vast majority of researchers pay attention to the evaluation of technological efficiency, ignoring the economic aspects of the proposed solutions. This study considered the dependence of the economic effect on the geological reservoir and technological well characteristics for variant 1—intelligent horizontal well (HW) completion using autonomous inflow control devices and variant 2—conventional horizontal well completion using the open hole. Calculations of production levels and dynamics in the two variants were performed on a created sector hydrodynamic model of a horizontal well operating in the depletion mode. The analysis of the obtained results allowed us to determine the applicability criteria of the proposed configuration of formation and well characteristics at the object of study, as well as to establish general dependencies of the net discounted income of an intelligent well. As a result of this study, it was determined that the economic efficiency of intelligent well completion with the use of autonomous inflow control devices relative to conventional well completion increases with decreasing permeability and drawdown pressure on the reservoir and reaches maximum values at the object of study at the thickness of the oil-saturated part of the reservoir about 5–6 m and the location of the wellbore in it at 35–40% of the thickness of the oil-saturated part below the gas–oil contact (GOC). This article covers the research gap in evaluating the economic efficiency of intelligent HW completion using AICD relative to conventional HW completion in oil rims.

1. Introduction

According to statistics from the U.S. Energy Information Administration, 81% of wells drilled in the U.S. in 2021 are horizontal or directional. In the Russian Federation, the share of horizontal drilling to production drilling reached 45% as early as 2019 [1]. One of the known features of horizontal wells is the heel–toe effect, which is a pressure drop along the wellbore due to friction. Such an effect creates an uneven drawdown pressure on the reservoir, which can lead to an early breakthrough in the formation of water or gas. This effect is intensified by formation heterogeneity, variable distance between the wellbore and fluid contacts, and inhomogeneity of reservoir pressure in different areas of the formation.

Various methods are used to combat early breakthroughs of unwanted fluids in horizontal wells. These include the following [2]:

-

Limiting the flow rate/drawdown pressure to the reservoir;

-

Intermittent well operation;

-

Regulated well operation;

-

Installation of inflow control devices.

The main difference between inflow control and other methods of preventing premature water or gas breakthroughs is that it regulates production by changing the inflow to the wellbore rather than by changing the operating mode of the well. This feature greatly expands the scope of application of inflow control devices (ICDs), as they can be applied to fields that are considered uneconomic to operate with older methods of controlling unwanted flow. Inflow control devices are categorized into the following [3]:

-

Passive inflow control devices (ICDs);

-

Autonomous passive control devices (AICDs);

-

Active (interval control valves, ICVs);

-

Autonomous inflow control valves (AICVs).

Passive ICDs are selectively installed filter elements that are placed on well intervals that are prone to early water or gas breakthrough. They do not have a control function, but they do help equalize the distribution of flow throughout the wellbore, thereby eliminating the negative effects created by uneven pressure distribution.

AICDs are a more advanced form of passive ICDs and rely on the dynamic properties of the fluid as it flows through the device. AICDs respond to fluid flow differentially through their design, which includes a specific shape of nozzle or valve that prevents the flow of lighter/heavier fluids [4].

AICVs rely on differences in the viscosity of reservoir fluids. In a typical AICV design, there are two conduits, one of which is an auxiliary conduit used to control the valve. A portion of the flowing fluid loses energy through friction as it passes through the auxiliary channel and is then released into the space below the valve. If the fluid has a high enough viscosity and has lost enough energy, the pressure on the valve created by it will not be enough to close the valve. However, if the fluid has low viscosity, it will have lost little or no energy and will begin to exert pressure on the valve, causing it to close.

Active ICDs are the only type of ICDs for which it is possible to change their characteristics after installation. Remote control of active ICDs is accomplished through hydraulic or electrical control communications. One of the simplest active flow control device designs is the sliding sleeve door (SSD), which allows for the selective shutoff of flow to a specific part of the well.

At present, a huge number of different designs of AICDs have been created and are being developed. A number of researchers distinguish two main types of designs: those with the presence of moving parts and those without them [5]. AICDs with moving elements (floating disks) are 2–3 times more effective than passive ICDs, but the presence of moving elements makes them vulnerable to the presence of mechanical impurities and breakdowns. AICDs without moving elements are more reliable and resistant to mechanical impurities. They utilize a dual inlet design with a funnel equipped with baffles. Due to the higher viscosity of the oil, it travels a shorter path to the outlet nozzle than water, which has to overcome much more resistance to enter the well due to the design of the device. Depending on the design of the device, it is possible to have a configuration in which the viscosity separation of the fluids is accomplished at the funnel inlet in a Y-shaped channel. AICDs with a Y-channel inlet are commonly referred to as Fluidic Diodes, and those without a Y-channel inlet are referred to as Phase Selection Controllers.

There are also known designs of AICDs designed to restrict the flow of liquid in gas wells. Their principle of operation is that at the inlet to the device is installed a spring-loaded ball, which is able to close the channel leading into the well in case of penetration of water or oil. This design is very common, and when the spring-loaded ball is placed on the other side of the inlet channel, it allows for limited gas inflow (analogous to the “floating disk” design, but with a ball instead of a disk).

The number of different applications of smart wells is gradually increasing, and many researchers predict a steady increase in the number of smart wells in the future [6]. Successful applications of smart wells have been reported in multilayer production, gas elevator automation, water and gas injection, deepwater well operation, heavy oil production, small reservoir operation, multilateral wells, and low-yield reservoir development [7].

All applied intelligent technologies in oil and gas wells can be divided into two categories: intelligent data algorithms and intelligent equipment and technologies [8].

As a result of analyzing scientific publications on the topic of using information and communication technologies (ICTs) in intelligent field development control and management systems, four main categories that most fully reflect various aspects of working with data have been identified [9]:

• Analyzing and storing big data (BD);
• Industrial internet of things (IIoT or IoT) applications;
• Application of digital twins (DTs);
• Use of artificial intelligence and machine learning (AI, ML).

The main problem is the lack of knowledge in the technical and economic evaluation and justification of the application of intelligent horizontal well (HW) completion using AICD compared with conventional HW completion in oil rims. The low level of application of intelligent completion using AICD is due to the lack of experimental data on the implementation of such projects. Therefore, the main task of this research was to provide practical substantiation of the variants of intelligent completion of hydrocarbons, which we modeled for the first time for specific conditions of an oil deposit and proved that under certain reservoir and well characteristics, it is most economically efficient to apply intelligent completion of hydrocarbons using AICD. This fact has practical significance and can be used in project planning and in the evaluation of proposed technological solutions.

This paper studies the main dependencies of the applicability of intelligent completion using AICD based on economic criteria. Due to the lack of possibility of full-fledged modeling of intelligent data acquisition and analysis systems, this study presents a description of the basic principles and limitations of the application of the corresponding systems in intelligent fields.

The purpose of this work was to identify the dependence and optimization of technological and economic criteria of applicability of the technology of intelligent wells using autonomous inflow control devices.

The objectives of this study include the following:

• Analyzing the scientific literature in the field of smart wells and inflow control devices for the purpose of scientific background review;
• Designing a hydrodynamic model based on selected geological and physical reservoir characteristics in tNavigator 22.2;
• Estimating the economic efficiency of intelligent well completion using autonomous inflow control devices in relation to conventional well completion based on the generated hydrodynamic model.

2. Literature Review

2.1. The Concept and Elements of an Intelligent Well

The term “intelligent” describes a well equipped with sensors and telemetry systems capable of acquiring various information about well performance and controlling its operation in real time. The information and measurement systems used provide and increase the reliability of information and operational systems for production monitoring and reservoir management, ensure project profitability, and increase the oil recovery factor. For the purpose of dividing intelligent wells according to their level of complexity and the devices used in them, scientists and organizations offer various classifications, separating wells according to both the structure of the wellbore profile and the type of production control [6,10].

The concept of intelligent completion refers to any type of downhole monitoring and/or remote control system capable of collecting, transmitting, and analyzing reservoir, production, and completion data while enabling remote control of the well, reservoir, and production processes [11]. Intelligent completion does not always involve an automated, self-regulated system and may rely on a manual interface to send settings to the well equipment [12].

To maximize control of both the intelligent well completion and the field development system as a whole, timely, high-quality production information about the status of the development process is required. In intelligent well systems, this information is collected continuously and transmitted in real time through a variety of downhole sensors. Typically, such systems are implemented as follows [7]:

• Detection and information acquisition system (downhole and other sensors that transmit information on key process indicators);
• Production control system (downhole tools with wireless control);
• Information transmission system (cable, fiber optic or wireless);
• Well data analysis system and numerical modeling and prediction technologies.

For the most efficient operation of the data acquisition and handling system, it is necessary to know as much information as possible about the future well at the design stage, as well as to have a plan for its management and to recognize in advance the problems it will face in the future [13].

Modern wired connection of downhole sensors allows them to be placed at a depth of more than 3800 m with data transmission speeds up to 1 Kbps [14,15]. The data received from the sensors are sent from the well cluster to the local control room using a radio modem [16]. The received data can then be analyzed and processed using an intelligent hydrocarbon field development management system. This system consists of a number of modules that provide automated analysis and optimization of the entire production complex up to the support of strategic decisions [17,18]. Effective functioning of such systems is virtually impossible without the use of modern information and communication technologies, such as IIoT and BD [19,20,21].

2.2. Principles of Operation and Limitations of Modern Information and Communication Technologies in Oil and Gas Field Development

The term “big data” is often used to describe databases of 1 petabyte or more. Analyzing such volumes of data allows us to extract patterns and relationships that can be used in making various decisions. There are six basic requirements for working with big data: volume, velocity of exchange, variety, value, ability to change, and veracity [22,23].

In the field of hydrocarbon production, the prospects of using big data analysis and storage technology are increasing because of the latest developments in seismic surveys, monitoring fluid behavior in reservoirs, more accurate measurement devices, and so on [23,24].

The major challenge of implementing database analysis is the amount of capital investment associated with data recording, storage, and analysis, as well as the low efficiency in the absence of necessary infrastructure. Implementation of database storage and analysis requires efforts to ensure the efficient utilization of available software and computing power, functionality, information security, and maintenance [25]. Among the organizational challenges in implementing a database system, there are interactions between different departments to ensure efficient workflow, standardization issues, data confidentiality, data ownership, and intellectual property rights [24,26]. Summarizing all the difficulties and limitations in the application of database analysis and storage technology, we can conclude that the main problems of implementation and effective use of the technology are the following [27]:

• Insufficient infrastructure readiness;
• Shortage of trained professionals;
• Financial constraints;
• Concerns about data security.

Internet of things (IoT) technology is a system for collecting and transmitting data both between different devices and in operator-device systems. It is capable of realizing an efficient data transmission system between people and objects, used to monitor and intelligently control the processes of each connected subsystem, which increases productivity and production safety [28]. By combining IoT with BD technologies, it is possible to achieve an almost instantaneous exchange of production data, which can significantly optimize the traditional manufacturing process [29].

IoT is a heterogeneous platform consisting of both devices and software. The heterogeneity of hardware and software parts brings to the discussion the problems of standardization and compatibility [30]. However, these problems are local and are solved by a reasonable combination of hardware and software parts of the system. The global problem in the application of IoT systems is the problem of information security [31]. The smallest flaw in the IoT security system can lead to the hacking of control systems, the damage from which can amount to billions of dollars.

Digital twin technology has many definitions, summarizing which we can conclude that a digital twin is a computer system that collects information about a physical object or a system of objects as input data and creates, on the basis of process simulations and data analysis, output information used to make various decisions [32,33].

In the development and operation of oil and gas fields, DT is applied both to predict various states of the real system and to control it based on the data obtained from simulations. Thus, the basic principle of DT in oil and gas field development and operation is two-way data fusion and flexible application of fractal logic [34].

Among the problems most frequently encountered in the application of digital twins, researchers highlight the following [32,35]: scoping and directional selection; lack of standardization; information security; data ownership issues of data sharing; accuracy and reliability of results; functionality of digital twins; application of oil and gas industry experts experience in the development of digital twins; problems associated with existing business models, conservative personnel, and company policies; storage and analytics; maintenance of digital twins; downplaying the role of digital twins in decision making.

Machine learning (ML) techniques are a group of methods often used in artificial intelligence that predict new properties of data based on known properties discovered from training data [36].

Artificial intelligence and machine learning technologies are used in the oil and gas industry for the following [37,38,39]:

-

Field history matching;

-

Forecasting production dynamics;

-

Development project optimization;

-

Analysis of enhanced oil recovery methods;

-

Identification of reservoir structural elements;

-

Prediction of various properties during drilling operations.

Various applications of artificial intelligence are capable of building preventive repair and maintenance systems for various equipment, drawing conclusions about the causes of production incidents based on the analysis of a sample of data or predicting the values of technological indicators [40,41,42].

Machine learning and deep learning are currently the dominant approaches used in the application of artificial intelligence throughout the oil and gas industry [37]. However, machine learning and deep learning algorithms are “black boxes”—there is no obvious correlation describing why systems based on them make the judgment that is the output value.

Artificial intelligence tools require high-quality information in a reasonable amount to function properly. Hence, access to quality information is a major empowering factor in the application of artificial intelligence [43]. Oil and gas fields produce large amounts of information, the veracity of which cannot always be ensured. Considering all the circumstances of AI technology applications, their main problem is to provide the necessary data. That is why, despite all the advantages, there is a limited demand for AI technologies, which can also be explained by the lack of necessary competencies of oil and gas industry specialists, threats of external cyberattacks, and questions about data ownership [44].

Because of the uniqueness of digitalization projects in terms of cost-effectiveness evaluation, most researchers have come to the conclusion that some system of integrated assessment of the most significant indicators is necessary to provide a systematic approach to management decision-making. In addition to NPV, IRR, ROI, PI, and DPP, they usually calculate a separate integrated indicator that characterizes the benefits from the application of digitalization technologies that cannot be directly evaluated economically [45,46].

2.3. Applicability Criteria and Experience of Application of Autonomous Inflow Control Devices

In the oil and gas industry, there is a strong belief that ICDs can effectively limit the inflow of undesirable fluids under various conditions, including in combination with enhanced oil recovery techniques [47,48]. The development of inflow control technologies also leads to the intellectualization of displacing fluid injection processes, but the use of inflow control devices is of a point character and does not bring super results [14].

Many domestic and foreign researchers have noted the positive experience of the application of intelligent completion technologies with the installation of inflow control devices to solve the problem of early formation water breakthrough [49,50,51]. In Russia, the use of AICD in various capacities has been applied in a number of fields: Tazovskoye, Srednebotuobinskoye, Zapadno-Messoyakhskoye, Yuri Korchagin’s, and Vankorskoye, Tagulskoye. The design and principle of operation of the AICD are shown in Figure 1.

The Srednebotuobinskoye oil and gas condensate field has had a positive experience with the AICD application. As of the end of 2022, 12 intelligent completion configurations have been installed at the field, which showed a more than twofold increase in oil production in horizontal wells and a more than twofold decrease in the gas factor (GF) in multilateral wells [52]. Also, the positive influence of AICD, with an increase in liquid flow rate of 1.5–2.5 times and a decrease in GF, was observed in the field of Eastern Siberia [53].

Foreign subsoil users have repeatedly tested technologies of intelligent well completion [54,55,56]. The conclusion about the expediency of intelligent well completion using AICD and SSD was received by foreign researchers on the basis of the experience of the application of AICD for GF reduction at well W-028 at one of the UAE fields [57].

Due to the varying effectiveness of AICD applications depending on field and reservoir conditions, it is critical to have an understanding of the optimal conditions. Determining such conditions is a complex optimization problem that requires many times more optimization iterations than conventional completions [58]. In the course of various computational experiments and pilot industrial works, researchers have derived a rather capacious list of conditions for the technological efficiency of using AICD. In particular, it has been established that the efficiency of watercutting begins to decrease sharply at an oil viscosity of fewer than 0.2 mPa·s and, in general, gradually decreases with decreasing oil density, which corresponds to the approximation of oil properties to water properties [59,60].

If the viscosity of oil is close to that of water, the pressure drop (flow resistance) created by the AICD becomes higher. As the proportion of oil in the inflow increases, the difference between the oil and water inflow resistance increases due to a decrease in the oil inflow resistance [60].

Watercuts also have a significant impact on the efficiency of AICD. The difference in water and oil inflow resistance between 0–40% watercut varies steadily, first decreasing and then increasing. For watercut between 40% and 80%, the efficiency of AICD increases significantly with the increase in water content. For watercut above 80%, the efficiency increase is insignificant but has dynamics higher than in the case of low water content. The difference in resistance to inflow depending on the watercut of production can reach 134% [60]. Product watercut affects completion efficiency much more significantly than reservoir fluid properties.

A number of studies and experimental calculations have shown that the greatest influence on the final performance of an AICD is its design, wellbore location, and applicability to specific reservoir conditions [61,62,63].

In general, the technological efficiency of different types of AICDs that rely on differentiated resistance generation for different phases of reservoir fluid varies along similar parameters. A study by Eric Broni-Bediako et al. investigated the technological efficiency of intelligent completions using active ICDs that shut off the inflow in the event of high watercut or gas factor [64,65]. In some ways, such an algorithm is similar to the principles of AICD operation, so the results of this study can also be attributed with some confidence to the study of AICD efficiency. As a result of this study, it was proved that ICDs show greater technological efficiency with increasing permeability of reservoirs [66]. This is because greater permeability increases the inflow rate and, hence, the cumulative production. Thus, the more fluids, particularly water and gas, flow into the ICD, the greater the pressure drop it generates, functioning more actively. Also, in oil rim development, the efficiency of the ICDs increases as porosity decreases [66].

Special attention should be paid to the rate of achievement of final indicators. Thus, higher permeability corresponds to faster hydrocarbon recovery, which loses in pace to more favorable conditions on final oil production only at the horizon of 5–10 years. This observation plays a key role in the question of the economic viability of AICD, as it is the early periods of the well that are critical for economic evaluation purposes.

According to the study by Eric Broni-Bediako et al., the improvement of bottomhole formation zone conditions (skin factors) leads to an increase in the efficiency of ICD. Moreover, the worse the skin factor, the earlier the value of the minimum current PI is reached, which then transitions to a steady increase [66]. In the study by Ali Moradi et al., special attention was paid to the analysis of the technological efficiency of AICD depending on permeability parameters. The study proved that formation anisotropy in terms of permeability has the least influence on well performance, second to absolute permeability and especially relative permeability [67]. At the same time, there is a tendency for the sensitivity to variation of these factors to decrease in wells with more advanced inflow control (in the series of openhole-passive ICD-AICD AICV).

In a study by Austin Afuekwe and Kelani Bello, it was found that as the ratio of vertical to horizontal permeability increases, the oil recovery efficiency of wells equipped with an ICD decreases. It was observed that as the ratio becomes higher than 0.1, there is early vertical water breakthrough, causing the device to have to provide more resistance to inflow, thereby making it more difficult for oil to enter the well [11].

The Eric Broni-Bediako et al. study also considered variation in the location of the oil-water contact. As the thickness of the oil-saturated horizon decreases, as well as with its sharp increase, the effectiveness of inflow control devices also decreases [66]. This is due to the fact that in the case of large oil-saturated thickness, water control by means of ICDs is very effective; it leads to insignificant water inflow and additional inflow resistance for oil, worsening the oil displacement process. In the case of small oil-saturated thickness, the ICD is unable to provide sufficient water inflow resistance and experiences water cone breakthrough within a few months, resulting in no water control and increased resistance to inflow, which leads to a decrease in the total well flow rate.

The results of the Solovyov T.V. studies indicate that the use of AICD is technologically effective for gas breakthroughs and, if the viscosity of the oil is more than 1.5 cP, for water breakthroughs [68]. Also, one of the derived criteria is effective choke restriction of the unwanted phase (delta P (inflow resistance) for water/gas is 25%+ higher than for oil). In general, the prerequisites for AICD application almost overlap with the prerequisites for the breakthrough of undesirable fluids [68]:

-

Shallow oil rim thickness;

-

Permeability contrast along the borehole (5–10 times and more);

-

Separation of channel/floodland facies;

-

High-viscosity oil;

-

Horizontal well drilling close to oil–water (OWC) and gas–oil contact (GOC);

-

High vertical permeability.

Evaluating all the above criteria of applicability and efficiency of intelligent completion using AICD, we can conclude that the vast majority of researchers pay attention only to the technological efficiency of this technology, almost completely ignoring the economic efficiency. To solve this problem, this paper evaluates the efficiency of intelligent completions using AICD relative to conventional completions by the economic criterion.

3. Materials and Methods

This study used open Internet sources and articles by Russian and foreign scientists devoted to theoretical and practical issues of application of intelligent well completion using AICD to improve the efficiency of field development. The research scheme is presented in Figure 2.

The literature review includes consideration of the peculiarities of the technologies used in the operation of intelligent wells, as well as a review of the experience and applicability criteria of AICD. The analysis of this information, in conjunction with the study of the problems of economic evaluation of digitalization projects, allows us to systematize the knowledge in the field of application of intelligent well completion using AICD and identify the features that allow us to judge the insufficient representation of the criteria of economic efficiency of application of such well layouts.

The object of this study is a sector hydrodynamic model approximated by its characteristics of the Vankor oil-gas condensate field. This hydrodynamic model has geological and physical characteristics presented in

Table A1. Geological and physical characteristics of the modeled reservoir section (compiled by the authors on the basis of [96,97]).

No. in Order	Parameter	Unit of Measurement	Productive Formation: Modeled Reservoir Section
1.	Average roof depth	m	1899
2.	Absolute mark of OWC	m	−1556
3.	Absolute mark of GOC	m	−1550
4.	Absolute mark of GWC	m	-
5.	Reservoir type	-	gas and oil
6.	Collector type	-	-
7.	Oil and gas bearing area	thousand m2	9120
8.	Average total thickness	m	32
9.	Average effective oil-saturated thickness	m	6.097
10.	Average effective gas-saturated thickness	m	3.806
11.	Average effective water-saturated thickness	m	22.582
12.	Porosity coefficient	fractions of units	0.15
13.	Oil saturation factor of pure oil zone	fractions of units	0.95
14.	Oil saturation factor of oil–water zone	fractions of units	-
15.	Oil saturation factor of the reservoir	fractions of units	0.181
16.	Gas saturation factor of the reservoir	fractions of units	0.113
17.	Permeability	×10−3 μm2	87.499
18.	Net sand coefficient	fractions of units	0.93
19.	Stratification factor	units	3
20.	Initial reservoir temperature	°C	-
21.	Initial reservoir pressure	MPa	14.391
22.	Oil viscosity in reservoir conditions	mPa·s	7.476
23.	Oil density in reservoir conditions	g/cm3	701.545
24.	Oil density in surface conditions	g/cm3	864.236
25.	Oil volume factor	fractions of units	1.172
26.	Sulfur content in oil	%	-
27.	Paraffin content in oil	%	-
28.	Bubble point pressure	MPa	-
29.	Gas content	m3/tonne	142.952
30.	Dew point pressure	MPa	-
31.	Condensate density at standard conditions	g/cm3	-
32.	Condensate viscosity at standard conditions	mPa·s	-
33.	Potential content of stable condensate in gas (C5+)	g/m3	-
34.	Hydrogen sulfide content	%	-
35.	Gas viscosity under reservoir conditions	mPa·s	0.019
36.	Gas density in reservoir conditions	kg/m3	151.712
37.	Real gas factor (z-factor)	fractions of units	-
38.	Water viscosity in reservoir conditions	mPa·s	4.0
39.	Water density in surface conditions	g/cm3	1280
40.	Compressibility of
41.	Oil	1×MPa×10−4	2.639
42.	Water	1×MPa×10−4	2.558
43.	Rock	1×MPa×10−4	6.550
44.	Displacement factor (by water)	fractions of units	-
45.	Displacement factor (by gas)	fractions of units	-
46.	Productivity factor	m3/day·MPa	-
47.	Filtration resistance coefficients:
48.	A	MPa2/(thousand m3/day)	-
49.	B	MPa2/(thousand m3/day)2	-

of the Appendix A. Due to the size of the sector hydrodynamic model, this study has to be limited to the calculation of one horizontal well in the depletion mode (

Table A2. Basic information on the type and dimensions of the sector hydrodynamic model (compiled by the authors).

Parameter, Unit of Measurement	Parameter Value
Model type	Multi-segment, Black oil
Number of blocks	78,240
Model size in blocks	24 × 38 × 20
Block size, m	100 × 100 × 1.8–100 × 100 × 0.9
Number of local grid refinements	1 (25 × 10 main blocks (2500 × 1000 m) around the wellbore)
Block size at local grid refinements, m	50 × 50 × 0.6–50 × 50 × 0.3

of the Appendix A).

Comparable well layouts are shown in

Table 1. Comparison of well layouts with intelligent and non-intelligent completions (compiled by the authors).

Layout Element	Intelligent Completion	Non-Intelligent (Conventional) Completion
	Position along the Wellbore, m
Casing string 116 mm	4050	4050
Tubing string 76 mm	3840	2440
Packer	2400	2400
	2700
	3000
	3500
Perforations	2450–4050	2450–4050
Tubing plug	3840	Absent
Autonomous inflow control devices (AICD)	2600	Absent
	2900
	3300
	3700

. Intelligent completion divides the horizontal wellbore into several isolated inflow zones, which are controlled by separate devices in order to produce hydrocarbon reserves in the most efficient manner. The AICD settings are set as standard in accordance with the software training material.

The main characteristics of the modeled well are presented in

Table 2. Main characteristics of the modeled well (compiled by the authors).

No. in Order	Characteristic	Well Data
1.	Design depth, m:
	- Vertically;	1907
	- Along the borehole.	4050
2.	Type of well	horizontal
3.	Profile type	five-interval
4.	Maximum deflection angle, deg	30
5.	Depth of the productive (base) formation roof, m	1903
6.	Deviation from the vertical of the point of entry into the roof of the productive (base) formation, m	988

.

For the purposes of economic evaluation, a simplified financial and economic model was formed to estimate the most significant cost items. As a result of the well construction capital expenditures assessment, the result amounted to RUB 345,182,183 for conventional completion and RUB 399,301,534 for intelligent completion.

Under current policy conditions, it is assumed that all associated petroleum gas is delivered to customers in the Russian Federation at the average unified gas supply system price for combustible natural gas of RUB 4685 per 1000 cubic meters of gas [69]. In accordance with the current tax legislation of the Russian Federation, associated petroleum gas is not subject to mineral resource extraction tax [70,71,72]. The degree of associated petroleum gas utilization is assumed to be 95% [73,74,75].

The average price level of Urals oil in US dollars per barrel for the period from January to December 2023, calculated in accordance with Section 3 of Article 342 of the Tax Code of the Russian Federation, amounted to USD 63.07. The average value of the USD to RUB exchange rate for the period from January to December 2023, set by the Central Bank of the Russian Federation, amounted to RUB 84.66 per USD. All crude oil is planned to be sold in the Russian Federation. Based on this information, the realized price of crude oil is RUB 38,230.98 per ton. The calculated value of mineral extraction tax for the field under study was RUB 23,943.52 per ton [76].

The oil production well belongs to the fifth depreciation group, and its amortization period is from 7 to 10 years inclusive. Within the framework of this work, we will take the depreciation period of capital expenditures for well construction as 10 years. The property tax rate for oil production wells is 2.2% of the residual value of the property [77]. The income tax rate is 20%. Operating costs and estimates for well construction for two types of completion are based on open sources and are shown in

Table A3. Estimated operating costs of a well (compiled by the authors on the basis of [85,98,99,100,101]).

Cost Item	Unit of Measurement	Value
Operating costs of oil and gas production:
- oil and gas field gathering and transportation expenses
including semi-variable costs	RUB/ton of liquid	19.97
fixed costs	thousand RUB/exploitation well per year	469.59
- oil treatment expenses
including semi-variable costs	RUB/ton of oil	199.43
fixed costs	thousand RUB/exploitation well per year	740.04
- equipment maintenance and operation expenses
including semi-variable costs	RUB/ton of liquid	-
fixed costs	thousand RUB/exploitation well per year	3552.91
- shop expenses of oil production shops	thousand RUB/exploitation well per year	919.34
- costs of oil transportation to the point of sale	RUB/ton of oil (transportation via main oil transportation pipeline from points Aprelskaya, Vatyegan, Purpe to Ust-Luga oil depot)	3094.15
- costs of gas transportation to the point of sale	RUB per 1000 cubic meters of gas (transportation per 1000 km including main gas pipeline usage fee)	782.52

and

Table A4. Estimated well construction costs for two types of completions (compiled by the authors on the basis of [10,102,103,104,105]).

No. in Order	Element	Cost	Costs of Non-Intelligent Completion, RUB	Costs of Intelligent Completion, RUB
1.1	Site development	RUB 728,947	728,947	728,947
1.2	Disassembly of pipelines	RUB 122,323	122,323	122,323
2.1	Construction and installation of a derrick	RUB 11,522,853	11,522,853	11,522,853
2.2	Disassembly and dismantling of the derrick	RUB 1,774,827	1,774,827	1,774,827
2.3	Installation of well testing equipment	RUB 423,018	423,018	423,018
2.4	Dismantling of well testing equipment	RUB 34,808	34,808	34,808
3.1	Well drilling	42,456 RUB/running meter	171,946,800	171,946,800
3.2	Well cementing	4,085,42,456 RUB/running meter	16,554,250	16,554,250
3.3	Costs of the drilling team in the process of penetrating the pay zone	1,215,406 RUB/reservoir	1,215,406	1,215,406
4.1	Testing during drilling	RUB 2,689,952	2,689,952	2,689,952
4.2	Completed well testing	RUB 3,186,975	3,186,975	3,186,975
4.3	Wellhead equipment	depending on the completion	255,720	255,720 + 1,005,364 (state-of-the-art wireless telemetry system)
4.4	Bottomhole equipment	depending on the completion	11,148,000 (slotted liner)	23,106,088 (sensors) + 23,000,000 (AICD)
5.1	Field geophysical survey	11% of No. 3.1–4.4	24,376,427	28,332,406
6.1	Preparatory works for construction and installation works in wintertime	RUB 664,793	664,793	664,793
6.2	Operation of heating boiler plant	49,157 RUB/day	7,815,963	7,815,963
6.3	Amortization of cementing units	954 RUB/day	151,686	151,686
7.1	Overhead costs	19.7% of No. 1.1–6.3	50,158,711	58,022,839
8.1	Planned accumulation	5% of No. 1.1–6.3	12,730,637	14,726,609
9.1	Bonus supplements	2.66% of No. п.1.1–8.1	8,445,556	9,769,691
9.2	Bonus for work under a rotation system	0.53% of No. 1.1–8.1	1,682,761	1,946,593
9.3	Northern benefits	2.98% of No. 1.1–8.1	9,461,562	10,944,993
10.1	Contingency reserve	2.4% of No. 1.1–9.3	8,090,207	9,358,630
Total:			RUB 345,182,183	RUB 399,301,534

of the Appendix A.

In the course of the economic evaluation, the main technical and economic indicators (NPV, PI, PP, DPP, IRR) were calculated [78]. The discount rate was assumed to be 10% in accordance with the requirements of the rules for the preparation of technical projects for the development of hydrocarbon fields. In the course of the evaluation, the main efficiency indicator was NPV. For it, as well as for the difference in NPV between intelligent and conventional completion, dependencies on the reservoir and well characteristics are built, reflecting the economic efficiency of the application of intelligent completion with the use of AICD relative to conventional completion in different conditions [79,80].

NPV was calculated as follows:

$$NPV={\displaystyle {\sum }_{t=0}^n}{\displaystyle \frac{S_t}{{(1+r)}^t}}-{\displaystyle \frac{I_t}{{(1+r)}^t}},$$

(1)

where the following is true:

• $S_t$—cash inflow;
• $I_t$—cash outflow;
• r—discount rate;
• t—number of time periods (year from the beginning of project implementation);
• n—number of time periods (project implementation period in years).

The difference in NPV for intelligent and conventional completion was calculated as follows:

$$∆NPV={\displaystyle {\sum }_{t=0}^n}{NPV}_{IW t}-{NPV}_{CW t},$$

(2)

where the following is true:

• ${NPV}_{IW t}$—net discounted income from intelligent completion at the time period t;
• ${NPV}_{CW t}$—net discounted income from conventional completion at the time of time period t.

4. Results

Table 3. Variable values of indicators used in hydrodynamic calculations (compiled by the authors).

Indicator	Minimum Value	Intermediate Value 1	Average Value	Intermediate Value 2	Maximum Value
Average permeability, mD	29.166	-	87.499	-	291.663
Drawdown pressure, MPa	0.5	-	3.75	-	7.5
Thickness of the oil-saturated part of the reservoir, m	4	5 *	6	7 *	8
Location of HW borehole in relation to the oil-saturated part of the formation, % of depth	25	37.5	50	62.5	75

* Calculation is performed only for intelligent completion.

presents varying values of indicators used in hydrodynamic calculations. The table is based on the data of the modeled reservoir, where the maximum values of permeability and reservoir depression correspond to the average permeability in the reservoir and drawdown pressure limit. One variable parameter was selected for each calculation, and all other variable parameters in this calculation took the average value. As a result, 24 calculations were carried out.

A cross-section of the formation along the wellbore is shown in Figure 3.

The NPV estimation results for all the experiments conducted are shown in Figure 4.

Evaluating the results, we can conclude that the optimal conditions for the application of intelligent completion with the use of AICD at the investigated object in relation to conventional completion are minimum drawdown pressure, minimum permeability, thickness of the oil-saturated part of the reservoir of about 5–6 m and position of the HW in the reservoir at about 40–50% of the depth. It should be understood that these results were obtained on standard recommended settings of AICD, and the obtained dependencies may differ in their appearance depending on the used AICD, but during the whole time of calculations, the obtained dependencies kept their shape. Thus, for example, as the AICD resistance decreased, the NPV line of the smart completion began to approach the line of the conventional completion but still retained its shape on the ΔNPV graph, albeit in a very vertically compressed form.

The obtained dependencies on drawdown pressure and permeability can be easily explained by the production dynamics using AICD. In the study “Application of intelligent well completion in optimizing oil production from oil rim reservoirs” [66], similar hydrodynamic calculations were performed on the model of intelligent completion using active ICDs. As a result of the calculation, it was found that due to the necessity of completion to limit the breakthrough of undesirable fluids into the well, the production dynamics in the first years were strongly reduced, while at low permeability, such an effect was much less pronounced. The results of this study are presented in Figure 5.

Since it is the moment of payment that is of utmost importance for economic evaluation, reduced production in the first stages significantly worsens the final NPV; thus, relative to conventional completion, intelligent completion is most economically feasible in conditions of limited inflow (low permeability and low drawdown pressure).

Calculation results obtained for reservoir size and HW position in the oil-saturated part of the reservoir are much more susceptible to the influence of the characteristics of the individual reservoir and cannot be taken as reliably as the calculation results for drawdown pressure and permeability. Nevertheless, as the thickness of the oil-saturated horizon decreases, so does the effectiveness of inflow control devices. Thus, for each reservoir–AICD configuration, there is an optimal thickness interval of the oil-saturated part of the reservoir. This is due to the fact that in the case of large oil-saturated thickness, control of undesirable fluids with the help of ICDs is very effective, which leads to an insignificant inflow rate and, as a consequence, worse oil displacement. In the case of a shallow oil-saturated thickness, the ICD is unable to provide sufficient resistance and experiences water cone breakthrough within a few months, thereby almost immediately ceasing to be useful and providing only parasitic resistance. The same logic can be applied to the location of HW relative to GOC and OWC.

In addition to the study on the sector hydrodynamic model, a calculation was carried out using an AICD optimized by its characteristics for a case with an average permeability of 87.499 mD and a thickness of the oil-saturated part of the formation of 5 m. In the calculations, the wellbore was placed 2.25 m below the GOC, and the variable parameter was the drawdown pressure. The calculation results are presented in Figure 6. The criterion of economic efficiency for intelligent completion relative to the conventional one (ΔNPV ≥ 0) began to be fulfilled only at drawdown pressure less than 0.8 MPa. Also, the sharp transition of the curve became pronounced due to the fact that after a certain point, the drawdown pressure growth practically ceases to influence the rate of formation energy usage and the rate of achieving the formation pressure limit.

At the same time, it should be noted that although intelligent completion allows you to obtain a higher NPV, it shows inferior PI, PP, and IRR indicators compared with traditional completion, which is due to a large amount of investment and relatively small (about a few percent) increase in NPV.

The economic analysis is based on comparing NPV results in two ways: (1) costs for non-intelligent completion and (2) costs for intelligent completion. The obtained deviations of ΔNPV indicate the efficiency of the second method. The economic effect of the second method is obtained due to the reduction of operating costs (OPEX) for the lifting of liquids. The cost reduction is due to more uniform oil recovery due to the ability of AICD to limit the flow of undesirable fluids into the well and thus equalize the subsidence pressure of the reservoir. Operating costs were calculated for different reservoir permeability, oil saturation and reservoir depth conditions. The economic efficiency of intelligent completion was obtained for an average permeability of 87.499 mD and an oil-saturated thickness of 5 m. The wellbore was located 2.25 m below the GOC.

5. Discussion

By assessing the results of this study, we can conclude that the limited interest of subsoil users in digitalization technologies is quite justified: investments in digitalization can, to some extent, be classified as venture capital investments, as they promise a relatively small increase in NPV (no more than 10%, usually less than 5%) in exchange for a fairly significant increase in investment. To overcome most of the obstacles of digitalization and its effective use in the oil and gas industry, it is necessary to create a detailed roadmap with general methodological recommendations. For many enterprises not yet familiar with digitalization, significant changes are also needed at the organizational level to reassess the use of the latest technology solutions by managers. In addition, to accelerate the digitalization process, it is necessary to increase investment in human capital through the development of modern digital education programs that allow quick trainining of specialists in the required digital competencies [81]. Similar trends in the development of digitalization can be traced in the Energy Strategy of the Russian Federation for the period until 2035.

It is worth realizing that there are quite a large number of possible intelligent systems that can find application in the oil and gas industry. Their classification is an open task, and many researchers suggest dividing such technologies by the level of their influence on the production process. In the paper “Digital twins: current and future directions of modeling in engineering dynamics applications”, for example, five levels of digital twin abilities are distinguished [82]. According to this concept, the key features that distinguish digital twins from various object state monitoring systems are the ability to predict, learn and control.

One of the main limitations of the conducted study is a very modest data sample and lack of modeling of data handling in the smart field. The significance of the obtained data is confirmed by similar results of production dynamics with the ICD obtained by other researchers. The lack of data modeling in the smart field is a rather serious drawback, as the digital transformation of production can bring a number of different benefits that can be assessed both qualitatively and quantitatively.

The used sector hydrodynamic model has a number of assumptions:

-

The deposit is new and is operated in depletion mode until the minimum allowable reservoir pressure of 7.7 MPa is reached;

-

Drawdown pressure is set as a limit on any point of the well, not on the average along the wellbore;

-

The location of AICDs in the wellbore is uniform because the studied reservoir does not have pronounced heterogeneities, so the result of using such completion in more structurally complex reservoirs may be better;

-

The well operates continuously, without stoppage;

-

Formation fluid losses are assumed to be zero.

One of the most promising approaches to the digitalization of the oil and gas industry could be the concept of a modern specialist that meets the requirements of Industry 4.0 [83,84]. According to this concept, a modern specialist in the oil and gas industry should be able to perform both their traditional job duties and tasks related to automated data analysis or the use of AI to make various decisions. To achieve this goal, IoT-connected smart sensors and actuators are proposed to be installed on most existing facilities and devices to enable remote, time-limited decision-making and real-time engagement. The emergence of big data as a result of the function of such systems will inevitably create the challenge of selecting and developing the necessary infrastructure [85]. It should be realized that the limited use of only one modern information and communication technology in itself is quite inefficient, so it is necessary to immediately think about a set of means. Big data, the software for work with which is continuously standardized and implemented in the oil and gas industry, has been proposed to be analyzed using a variety of methods, the most advanced of which are self-learning neural networks [86,87]. By such a system, all existing data can be subjected to any method of analysis and forecasting depending on the current need.

From the point of view of BD system security, the two most critical areas can be distinguished: identity and access management and data encryption. Identity and access management prevents unauthorized persons from gaining access to the system through the use of passwords and biometric data [88]. At the moment, various researchers have proposed methods of encrypting information during its storage in databases and transmission, which allows one to choose the best option for each specific case [89]. Nevertheless, it is worth noting that regardless of the number of proposed solutions in the field of cybersecurity, this area needs deep standardization and a legislative framework to simplify the processes associated with the management and development of data protection systems [90].

In addition to BD and IoT systems, the third major area of digitalization of oil and gas fields is the use of digital twins. These three areas complement each other and derive different benefits when interacting jointly with data stored and generated by their standard processes. The article “Designing Digital Twins” analyzed publications on the topic of digital twins and found that about 86% of the evaluated publications offer solutions to existing problems, with most solutions published in the form of basic models [91,92,93]. Therefore, it was suggested that a focus on specific areas is needed to increase the rigor of architectural models for digital twins. For training neural networks, the amount and reliability of data is critical. Some researchers have proposed to enhance the sharing of field data between companies to accelerate the development of AI in the oil and gas industry [94].

Transformations in the labor market accompanying the process of digitalization of the industry will be inextricably linked to job cuts, the emergence of new professions and changes in personnel competencies. In order to ensure a smooth transition to digital management in the oil and gas industry, it is necessary to develop and adopt comprehensive measures aimed at studying the possible consequences associated with the automation of production, as well as the adoption of programs to train specialists in the required skills to work with the digital systems. The role of companies in digital transformation should be active, as their organizational environment will be, first of all, to face the problem of providing specialists who meet modern requirements. At the same time, for the more successful application of modern information and communication technologies, organizations and society need social changes that are aimed at increasing the social acceptance of new technologies [95,96].

In order to stimulate innovative research and technology implementation, it is proposed to stimulate the creation and development of small and medium-sized oil and gas companies to form a competitive environment in the oil and gas sector, which will stimulate the development and application of new technologies. At the same time, the control of cardinal technological renewal of the industry should still be carried out at the federal level [42]. In this case, it will be possible to form priority areas of development, attract the necessary funding, and stimulate developers and subsoil users to create, implement, and replicate high-tech intellectual property objects.

To summarize, the main problem of digitalization in the oil and gas industry is the rejection of a comprehensive problem-oriented approach: digital solutions are used to strengthen weaknesses for marketing purposes or locally. To overcome the challenges of applying modern information and communication technologies in the oil and gas industry, companies need a detailed roadmap for conducting digital business transformation [45].

Further research is planned to investigate the mechanisms for optimizing intelligent completion using ICD for specific reservoir conditions. The results of the research can be applied in practice when assessing the applicability of AICD in certain conditions.

6. Conclusions

As a result of analyzing the scientific literature in the field of intelligent wells and inflow control devices, the main components of intelligent wells were identified, and a hydrodynamic model of the reservoir well was compiled based on the data of a real existing deposit. For the purposes of this study, hydrodynamic calculations were carried out for two types of well completion: conventional and intelligent well completion using AICD. On the basis of scientific publications and open sources, a financial and economic model was developed, and the results of the hydrodynamic calculations were evaluated economically.

As a result of analyzing the results of hydrodynamic calculations, the dependences of NPV on various reservoir and well characteristics were obtained for conventional and intelligent completion. Based on the data obtained, a conclusion was made about the applicability of AICD in various conditions, and the reasons for the limited application of intelligent completion technology in the oil and gas industry were substantiated. The main feature of using AICD for oil production turned out to be the unique production dynamics associated with limiting the inflow of undesirable fluids. Thus, completions using AICD can achieve higher oil recovery but at a much more modest rate, which dramatically degrades the technical and economic performance of this type of completion because the first, most economically important periods of well operation show lower production than conventional completions.