Integrating CO₂-EOR and Sequestration via Assisting Steam Huff and Puff in Offshore Heavy Oil Reservoirs with Bottom Water

Abstract

CO₂-assisted steam huff and puff is an effective method to improve oil recovery and store CO₂ in heavy oil reservoirs. However, few studies focused on complex geological formations, such as bottom water. The bottom water condition not only complicates the process of oil production and CO₂ sequestration, but also makes migration and distribution of oil, water and CO₂ unclear. In this paper, a numerical geological model of an offshore heavy oil reservoir with bottom water is established to analyze the influence of bottom water on injection and production parameters, oil recovery and CO₂ storage capability under vertical and horizontal well layouts. The results show that the bottom water could maintain the formation pressure, but reduce the steam chamber radius and heavy oil utilization area, increase water production and decrease the oil–water ratio. CO₂ could enhance oil recovery in the bottom water reservoir. Oil development indicators of the horizontal well are higher than the vertical well. Meanwhile, CO₂-assisted steam huff and puff use in the bottom water reservoir can create a high-pressure and -temperature environment to make CO₂ supercritical, as it has better CO₂ storage capability and efficiency. The CO₂ storage efficiency of the horizontal well is 63% larger than the vertical well. Thus, the horizontal well layout should be used as a priority if bottom water presents. Conducted analysis of bottom water formation sensitivity parameters shows that the advantageous formation conditions are high oil saturation, porosity of 0.2–0.4 and permeability of 2000–3000 mD. The influence degrees of each formation parameter were evaluated as well.

1. Introduction

1.1. Research Background

As a non-renewable resource, oil reserves have been declining rapidly in recent decades. With the global demand for crude oil continuing to rise, new strategies to enhance oil production are urgently needed. Heavy oil is widely recognized as an important energy resource, with global recoverable reserves of about 1.9 billion tons accounting for nearly 36% of global recoverable oil reserves, according to the International Energy Agency (IEA) statistics [1,2].

Over the past seven decades, China’s petroleum industry has undergone rapid development, evolving from an oil exporting country to a net oil importing country. The contradiction between oil supply and demand is increasingly acute, which has become a bottleneck restricting economic and social development [3]. As a rich resource in China, heavy oil is mainly distributed in Xinjiang, Liaohe, Bohai Bay and other oilfield blocks [4]. At present, the total proven reserves are about 8.2 billion tons, of which the proven reserves of onshore heavy oil reserves are about 4 billion tons, mainly distributed in Liaohe, Xinjiang, Shengli and other major blocks. The proven reserves of offshore heavy oil are about 4.2 billion tons, which are concentrated in the Bohai Bay area [5].

Heavy oil has great viscosity and poor mobility in the reservoir, due to the large amount of gum and asphaltene within. However, it has been observed previously that the viscosity of crude oil decreases sharply with an increasing temperature, consequently diluting heavy oil and improving its mobility in the reservoir [6,7,8,9]. Based on this principle, steam huff and puff, an in situ thermal recovery technology is developed, within which the main feature is allowing for three processes of steam injection, well soaking and production in the same well.

Nevertheless, only crude oil in a limited area around the wellbore can be produced through steam huff and puff. The oil production effect is declining with huff and puff carrying out and the oil recovery rate is generally only about 15%. Meanwhile, steam huff and puff will cause a high input cost, high energy consumption, high CO₂ emission intensity, etc. [10,11,12].

1.2. Development of Increasing Oil Recovery with CO₂

Over the past century, to address the challenge of declining productivity in the later stages of steam huff and puff and to further enhance the sweep efficiency of crude oil within the reservoir, the technique of co-injecting gas with steam during the later phases of steam huff and puff has been extensively studied. As a non-polar molecule, carbon dioxide is easily soluble in crude oil. Hence, a number of studies were on the production-enhancing effect of CO₂ on crude oil, as well as trial productions that were conducted in various oil and gas fields. In 1963, Welker [13] found that the solubility of CO₂ in crude oil is related to the relative density and viscosity of crude oil. The higher the density of heavy oil, the lower the solubility of CO₂. The higher the viscosity of heavy oil, the better the viscosity reduction effect of CO₂. Later, Monger et al. [14] observed that when the viscosity of CO₂ in the reservoir is reduced, the oil production-enhancing rate is increased. It is also found that CO₂ can extract the light hydrocarbon components in the crude oil, thereby increasing the production of a single well. Haskin et al. [15] carried out laboratory analyses of CO₂ huff and puff in different viscosities of crude oil and conducted a trial production at several oil fields in Texas. They believed that crude oil expansion, dissolution and viscosity reduction are the main factors to enhance oil recovery.

Further studies on the enhancing oil recovery (EOR) mechanism and controlling factors of CO₂ were carried out in succession. Yang et al. [16] and Wang et al. [17] discovered that CO₂ dissolved in crude oil can reduce the interfacial tension between oil and water and increase the formation pressure around the wellbore. During the production period, dissolved gas was generated to promote the migration of crude oil, thereby increasing the production of a single well. Pu et al. [18] found that it increased both the oil displacement efficiency and sweep efficiency, enhancing the oil recovery by introducing CO₂ into the steam flooding process. Zhang et al. [19] conducted laboratory experiments and numerical simulations and found that factors such as porosity and reservoir thickness have greater influence on the effect of CO₂ huff and puff, while the permeability and liquid production rate have less influence. Moreover, CO₂ huff and puff in the horizontal well has a more obvious stimulation effect than the vertical well. Chen et al. [20] pointed out that the contact differences in space and time between the CO₂ and the crude oil, as well as the scale of CO₂ mobilization for attic oil caused by gravity, are the main reasons for a better CO₂ huff and puff effect in the horizontal well than in the vertical well.

On the other hand, Or et al. [21] researched the CO₂ gas microbubble of foamy oil through numerical simulation and pointed out that the effect of the initial oil saturation and CO₂ dissolution zone are the controlling factors of heavy oil production. Zhou et al. [22] conducted further research and developed a dynamic model to match the foamy oil stability of the heavy oil–CO₂ system, with high agreement being achieved between the experimental data and the calculation results. To stabilize and reinforce CO₂ foam in foamy oil, Yang et al. [23] investigated the foam aging rules and CO₂ foam stability mechanism of synthesized acid-resistant hydrophobic polymer nanospheres, and obtained the best scheme of CO₂ foam reinforcing, which showed the highest foam stability.

Over the past decade, CCUS has become increasingly popular. Hill et al. [24] investigated and discussed the possibility, advantages and implementation conditions of CCUS by EOR technology, pointing out that EOR can effectively reduce carbon emission. Jin et al. [25] undertook experiments to investigate properties of Bakken shales and the process of scCO₂ extraction of shale oil. These experimental results demonstrate the potential of scCO₂ injection to enhance ultimate oil recovery while concurrently offering significant CO₂ storage potential within the Bakken Formation. Zhang et al. [26] conducted experiments of CO₂ immiscible huff and puff, revealed microscopic oil displacement characteristics and the evolution of gas saturation states, and provided a theoretical basis for CO₂ storage through immiscible huff and puff. Guo et al. [27] discovered that a high temperature is conducive to CO₂ storage but not conducive to displacement oil efficiency. Thus, the integration of CO₂ storage and oil recovery is feasible for strong edge water reservoirs.

In conclusion, it is now well established from a variety of studies that with the functions of viscosity reduction by dissolution, pressurization and formation of foamy oil with CO₂ can be used in miscible flooding, immiscible flooding, steam huff and puff, gas-assisted steam huff and puff, etc. [12]. Employing CO₂-assisted steam huff and puff represents a dual-purpose approach, integrating the synergistic effect to effectively enhance heavy oil recovery and achieve substantial CO₂ sequestration within the reservoir.

1.3. Previous Studies of Bottom Water Reservoir

Reservoir conditions are inherently diverse and complex, posing significant challenges during oilfield development. Numbers of studies, including those mentioned above, basically paid attention to reservoirs with little geological complexity: for example, the edge and bottom water conditions. Edge and bottom water are very common in global heavy oil reservoirs. The bottom water coning phenomenon poses significant and persistent challenges to reservoir development, having impacted oilfield operations for nearly a century, and remains a difficult issue to resolve even today [28,29,30,31]. A well-known example is the active edge and bottom water of the reservoir in Bohai Bay, China, where the edge and bottom water are present in approximately 57% of the discovered oilfields [32]. Abundant studies of heavy oil with bottom water have been undertaken for decades. In 1931, the research and analysis of Wright [28] on the production capacity of edge and bottom water reservoirs showed that the oil recovery efficiency was greatly affected by the energy of the edge and bottom water. Moreover, the rapid rate of bottom water intrusion into the reservoir also has a serious impact on the production effect of crude oil. Then, Muskat [33] analyzed the production behavior of the bottom water reservoir and found that the flooding efficiency of the bottom water reservoir can be obtained by using the parameters of reservoir thickness, well conductivity and ratio of vertical to horizontal permeability. In the 1990s, Saleh [34] extended prior modeling approaches to derive an analytical formula for water channeling velocity in bottom water reservoirs. Subsequently, Yu et al. [35] summarized the models developed by previous researchers and conducted a numerical simulation study to investigate water intrusion mechanisms during the production of sandstone reservoirs with bottom water. Their results identified several key factors controlling bottom water coning, including the oil-flow ratio, the vertical-to-horizontal permeability ratio, the distribution of interlayers and the oil production rate.

In the 21st century, the risk of bottom water coning has been systematically investigated. Zhou et al. [36] employed history-matching numerical simulations to analyze the dynamic characteristics of horizontal well flooding and their influencing factors. Their findings offer valuable insights for horizontal well planning, completion design, and late-stage production optimization. Using a synthetic fracture network model and numerical simulation, Lee et al. [37] investigated the water breakthrough phenomenon in fractured basement reservoirs with bottom water aquifers. Their study analyzed the effects of various factors, including the production rate and the presence or absence of capillary pressure in fractures, on water coning and breakthrough timing. Liu et al. [38] used numerical simulation methods to explore and analyze the location of horizontal well bottom water channeling and established the equation of bottom water cresting and reservoir’s water breakthrough time with gravity factors considered. They found that bottom water will preferentially invade from the middle of the horizontal well layout. By establishing and solving the pressure response analysis model, Sun et al. [39] analyzed the influence of the water volumetric multiple on the pressure response, and found that there are six main flow stages in horizontal well seepage in the finite bottom water reservoir.

In conclusion, the studies mentioned above analyzed the law, the migration characteristics and influence of bottom water on many methods of conventional oil and heavy oil recovery, which provide a technical basis and support for the oil production in bottom water reservoirs. However, these studies merely focused on conventional EOR methods, such as steam flooding, steam drive, water drive, and so on.

1.4. Research Purpose and Significance

As the application scope of CO₂-assisted steam huff and puff technology continues to expand, more challenging and complex geological conditions are encountered. So far, few studies in this field, including those referenced earlier, have paid attention to specify the influence of bottom water on the CO₂-assisted steam huff and puff process and CCUS capability, since the bottom water will seriously affect the oil recovery efficiency, formation water flow law and other parameters. As a result, correct CO₂-assisted injection and production parameters cannot be determined in the field production of heavy oil reservoirs with bottom water, thus resulting in the failure to implement the CO₂-assisted steam huff and puff stimulation process. Oilfield engineers may choose to abandon developing heavy oil reservoirs with bottom water if the development procedure is not clear, wasting the development potential of heavy oil production and CCUS capability.

Hence, this study not only focuses on the potential of CO₂-assisted steam huff and puff technology to increase oil production and the oil–water ratio, but also discusses the difference in the EOR effect of CO₂ and CO₂ storage efficiency with or without bottom water in the heavy oil reservoir. To this end, a typical offshore bottom water heavy oil reservoir model is built. Based on both horizontal and vertical well configurations, a comprehensive injection and production parameter optimization is first conducted to identify the most economically viable development scheme. Subsequently, the development indicators and parameter issues are compared and analyzed, aiming to explain the influence of bottom water on CO₂-assisted steam huff and puff process and evaluate the CO₂ storage efficiency in such an environment, providing technical reference and service guidance for practical production and CCUS projects.

This study is primarily a conceptual and theoretical validation. This positioning stems from the current scarcity of dedicated research and field applications specifically targeting the combination of heavy oil reservoirs with strong bottom water and CO₂-assisted steam huff and puff technology, making it difficult to obtain direct external field data for benchmark validation. The present work aims to establish a fundamental theoretical framework and conduct a feasibility analysis for this specific domain, thereby providing a reference for future development planning of such challenging reservoirs. We fully recognize the importance of external data verification for scientific rigor. Consequently, clear follow-up research plans have been established, which include both systematic physical simulation experiments to calibrate the numerical model and efforts to pursue collaboration on pilot field trials to acquire the necessary validation data.

2. Numerical Simulation

2.1. Geological Model

The offshore reservoir data used for numerical research and analysis in this paper are referenced from the research of Wang [40], who pointed out in his study that in the development process of heavy oil reservoirs, a conventional square well pattern is generally adopted. Considering the symmetry of the development well pattern, the maximum range that can be swept by a single production well can be selected as the simulation area. In the present study, a typical geological model of a heavy oil reservoir was established with a total of 12,500 grids, as shown in Figure 1a. The reservoir consists of oil–water two-phase fluid with an oil saturation of 0.7 and water saturation of 0.3 at the initial stage. More parameters of the geological model are shown in

Table 1. Geological model parameters.

Parameter	Unit	Value
Model size (length: width: height)	m	250:250:20
Gird size (length: width: height)	m	10:10:1
Buried depth of reservoir’s top	m	800
Initial pressure	kPa	8500
Initial temperature	°C	50
Porosity	/	0.3
Horizontal permeability	mD	2000
Vertical permeability	mD	1000
Initial oil saturation	/	0.7
Formation compressibility	kPa−1	8 × 10−6
Volumetric heat capacity	J/(m3 × °C)	2.45 × 106
Volumetric heat capacity of overburden and underburden	J/(m3 × °C)	7.87 × 105
Thermal conductivity of overburden and underburden	J/(m × day × °C)	1.496 × 105

. These parameters are derived from multiple sources and cover various oil fields such as the Bohai Bay Oilfield and the Liaohe Oilfield, aiming to form a typical bottom water heavy oil reservoir [40,41,42,43]. It should be noted that the capillary forces are not explicitly modeled in this study. This simplification is based on the following rationale: (1) The primary focus is on assessing the macroscopic performance of the physics of fluid flow in porous media, where viscous and gravity forces dominate the large-scale fluid displacement. (2) As supported by previous studies, in high-temperature steam environments and in contexts involving bottom water coning, the relative importance of capillary forces on the main flow pathways diminishes compared to viscous and gravity forces. The cited works suggest that the coning of mobile bottom water can overshadow the capillary effects at the reservoir scale under such conditions. While capillary forces can be crucial at the pore scale or in certain heterogeneous settings, their omission is a recognized simplification for the system-scale, mechanistic comparison intended here [37,44,45,46].

The models of vertical well and horizontal well layouts were established, respectively, as shown in Figure 1a. The vertical well is located in the center of the reservoir and perforated at all layers. The horizontal well is located in the center of the reservoir as well, buried 13 m from the top of the reservoir and 100 m in length.

A number of studies have shown that the water volumetric multiple can affect the dynamic change in pressure in the process of production. The larger the water volumetric multiple, the more abundant the energy of bottom water, and the more significant the impact on the oil recovery effect [39,47,48,49]. However, infinite bottom water can be more useful for highlighting the effect of bottom water on CO₂-assisted steam huff and puff, so it is chosen to use in the numerical simulation in this paper. The effect of infinite bottom water is achieved by implanting the analytical water body into the bottom of the reservoir model. The analytical water body adopts the Carter–Tracy [50] analytical model, which is suitable for water bodies with large thickness and can effectively simulate the water invasion phenomenon of infinite bottom water. However, according to the CMG-STARS simulator’s user manual, it is prone to iteration errors in the simulation process if the analytic water body has direct contact with the reservoir. So, adding an additional 10 m thickness of numerical water body between the reservoir and the analytic water body is necessary.

To sum up, two typical geological models with identical sizes of heavy oil reservoir were built, which are shown as Figure 1b, to compare the differences and advantages of the two well layouts in terms of enhancing oil recovery, providing a theoretical basis for subsequent studies. The only difference between these two models is the presence or absence of bottom water, in order to highlight the influence of bottom water in the follow-up study.

2.2. Fluid Components Numerical Model

A numerical model of fluid components was established in this paper, referring to the research of Cui et al. [51,52]. The physical property parameters of heavy oil used in the model were sourced from the Y oilfield in the Bohai Bay block, featuring high viscosity, low freezing point, low wax content, and high gum and asphaltene contents. The main fluid components are water, CO₂ and heavy oil in the numerical model.

To determine the flow behavior and phase saturations, the conservation equations in this reservoir, based on the principle of conservation of mass, are shown as follows:

$${\displaystyle \frac{\mathsf{\partial }({\rho }_i\varphi S_i)}{\mathsf{\partial }t}}+\nabla ·({\rho }_i{v}_i)=0$$

(1)

where ρ_i is the density of component i, kg/m³; Φ is porosity; S_i is the saturation of component i; t is time, s; and v_i is the velocity vector of component i, m/s.

The energy conversion and transfer process of the fluid in the reservoir are described by the energy conservation equation. Considering the existence of oil, water and gas in the reservoir, the energy conservation equation can be expressed in combination with the first law of thermodynamics and the characteristics of fluid flow:

$$\rho {\displaystyle \frac{dh}{dt}}+\nabla ·(\overrightarrow{v}h)=-{\displaystyle \frac{\mathsf{\partial }p}{\mathsf{\partial }t}}+\nabla ·(k\nabla T)$$

(2)

where h is specific enthalpy, J/kg; k is thermal conductivity, W/(m·K); and T is temperature, K.

The basic equations used to describe the flow of multiphase fluid in the porous reservoir are the multiphase Darcy formula (Equation (3)) and two-phase fractional flow, based on the research of Buckley and Leverett [53] (Equation (4)):

$${\overrightarrow{v}}_i=-{\displaystyle \frac{k_{ri}}{{\mu }_i}}+(\nabla p_i-{\rho }_{i}g)$$

(3)

$${({\displaystyle \frac{\mathsf{\partial }u}{\mathsf{\partial }t}})}_{S_D}={\displaystyle \frac{q}{\varphi A}}{({\displaystyle \frac{\mathsf{\partial }{f}_D}{\mathsf{\partial }{S}_D}})}_t$$

(4)

where k_ri is the relative permeability of component i; μ_i is the viscosity of component i; p_i is the pressure of component i; g is the gravitational acceleration, m/s²; and D means the displacing fluid.

The density and viscosity of heavy oil are 978 kg/m³ and 4502 cP in reservoir conditions. More physical parameters are shown in

Table 2. Physical parameters of heavy oil.

Parameter	Unit	Value
Density (ground)	kg/m3	993
Density (reservoir)	kg/m3	978
Viscosity (50 °C)	cP	4502
Viscosity (100 °C)	cP	387
Pour point	°C	12
Wax content	%	2.94
Asphaltene and gum content	%	38.29
Bubble point pressure	kPa	1128
GOR	m3/m3	9.55
Formation volume factor	/	1.015
Liquid compression coefficient	kPa−1	2.618 × 10−7
First coefficient of thermal expansion	°C−1	1.64 × 10−4

, some of which are based on the measurement data from the Y oilfield in the Bohai Bay block, while others are referenced from Cui’s article [51], since the latter uses the same crude oil sample. The viscosity–temperature curve of heavy oil (Figure 2), which references the field experimental data, has a great negative slope from 0 to 150 °C. The viscosity is only 119 cP at 150 °C, which represents a reduction of 97% from the viscosity at 50 °C. Such a difference demonstrates a peculiarity: that heavy oil’s viscosity reduces sharply with an increasing temperature, despite the former being quite heavy under the reservoir condition. Figure 3 shows the gas–liquid phase relative permeability curve and oil–water phase relative permeability curve used in this research, which reference the field experimental data in the Bohai Bay block, using the crude oil samples from the Y oilfield. It should be noticed that the oil–water phase relative permeability curve would shift to the right entirely, with the temperature increasing (gas–liquid phase relative permeability curve will not change with temperature in this model).

The dissolution equilibrium mechanism of the fluid component is represented by the K value function. The K value is a kind of phase equilibrium constant, which is related to the temperature and pressure. It can directly reflect the phase changes in fluids and dissolution results between different fluids:

$$K_i^{\mathrm{gw}}={\displaystyle \frac{y_i}{w_i}};K_i^{\mathrm{go}}={\displaystyle \frac{y_i}{x_i}};K_i^{\mathrm{ow}}={\displaystyle \frac{x_i}{w_i}}$$

(5)

where K^gw is the gas–water phase equilibrium constant; K^go is the gas–oil phase equilibrium constant; K^ow is the oil–water phase equilibrium constant; x_i is the mole fraction of component i in the oil phase; y_i is the mole fraction of component i in the gas phase; and w_i is the mole fraction of component i in the water phase.

The K value of a fluid component under certain conditions of temperature and pressure can be determined through the following:

$$K_i(p,T)=({\displaystyle \frac{K_{v1i}}{p}}+K_{v2i}·p+K_{v3i})·\exp ({\displaystyle \frac{K_{v4i}}{T-K_{v5i}}})$$

(6)

where K_V1, K_V2, K_V3, K_V4 and K_V5 are the first, second, third, fourth and fifth coefficients in the correlation for K value. p is the gas phase pressure.

The heavy oil in this numerical model is only oleic and no phase change will be occurred, so its K value coefficients are zero. Water can be aqueous or gaseous: hence, its K_V1 = 1.18 × 10⁷, K_V2 = K_V3 = 0, K_V4 = −3816.44, and K_V5 = −227.02. CO₂ can be gaseous or oleic since it can dissolve in water or oil, so the K value coefficients of CO₂ are set as 8.62 × 10⁸, 0, 0, 3103.39, and −272.99 for K_V1, K_V2, K_V3, K_V4, and K_V5, respectively. These K value coefficients were obtained from the operation manual of CMG software, based on the work of Reid et al. [54]. They can still accurately characterize the various fluid properties, even in a high-temperature environment with steam injection.

2.3. Approach and Process

In order to explore the influence of bottom water on the CO₂-assisted steam huff and puff process, several cycles of huff and puff were conducted on the model and the results were compared and analyzed. The specific steps are as follows: (1) The models with and without bottom water were established. Each model respectively contained two well layouts: a vertical well and a horizontal well. (2) A cycle of cold production was carried out on the models to simulate the initial development stage of the oilfield. The injection and production process parameters of the conventional steam huff and puff were optimized when the cold production was exhausted. Then, several cycles of conventional steam huff and puff were carried out. (3) After the production stimulation effect of conventional steam huff and puff was reduced, the injection and production process parameters of CO₂-assisted steam huff and puff were optimized and several cycles of CO₂-assisted steam huff and puff were undertaken. (4) The oilfield development indicators, remaining oil saturation/pressure/temperature distribution fields and CO₂ storage capability of the models with and without bottom water were compared and analyzed. The influence of bottom water on the CO₂-assisted steam huff and puff process was revealed and the CO₂ storage capability of the CO₂-assisted steam huff and puff technology in the bottom water reservoir was evaluated. (5) Furthermore, analysis of formation sensitivity parameters was conducted to clarify the influence changes and intensity levels of formation porosity, permeability and oil saturation on cumulative oil production, incremental oil recovery of CO₂ and CO₂ storage efficiency. The whole research process is shown in Figure 4.

3. Results and Discussion

3.1. Optimization of Injection and Production Parameters

A large number of research results show that the injection and production process parameters play an important role in the development of steam huff and puff [55,56,57]. The optimization method adopted in this paper is: (1) setting up a variety of parameter schemes to simulate; (2) comparing the simulation results, which are generally cyclic oil production, unless stated otherwise; (3) choosing the best scheme for a single parameter and (4) optimizing the next parameter in turn. In order to maximize the oil recovery of the horizontal and vertical wells, the injection and production parameters of the two well layouts were optimized, respectively.

Firstly, the conventional steam huff and puff process parameters of the no bottom water model were optimized. Five main parameters were optimized in the sequence: cyclic steam injection intensity (the explanation of injection intensity is shown in Appendix A), cyclic steam injection rate, soaking time, steam quality and steam temperature. It should be noted that the formation pressure can be effectively supplemented when the steam injection intensity of each cycle is changing successively, such as equal increases and variable increases. The cycle of conventional huff and puff can be extended with this method [58]. However, the conventional huff and puff stage is not the main content of this study. So, the method of constant steam injection intensity in each cycle for steam huff and puff was chosen to be used in this paper. During the conventional huff and puff, as well as the subsequent CO₂-assisted huff and puff stages, well production was stopped and the next cycle of huff and puff was entered if the daily oil production decreased to 1 ton/day.

After a number of cycles of conventional huff and puff, the production capacity was decreased. Several huff and puff cycles of later stages were selected to carry out one cycle of CO₂-assisted steam huff and puff to confirm the CO₂ injection timing by comparing the oil production of various schemes. Afterwards, the CO₂ injection intensity and CO₂ injection rate were optimized successively, with the same optimization method as the conventional huff and puff stage. The temperature of the injected CO₂ is 15 °C, representing gaseous CO₂ under surface conditions. The optimization results of the no bottom water condition are shown in

Table 3. Injection and production parameters of reservoir without bottom water.

Parameters	Vertical Well	Horizontal Well
Steam injection intensity (t/m)	150	40
Steam injection rate (t/d)	500	800
Soaking time (d)	30	60
Steam temperature (°C)	350	350
Steam quality	0.6	0.7
CO2 injection timing	Seventh cycle	Ninth cycle
CO2 injection intensity (m3/m)	7500	1500
CO2 injection rate (m3/d)	1500	50,000
Liquid production rate (t/d)	20	20
Minimum bottom hole pressure of production well (kPa)	200	200

. Every injection and production parameter was unchanged in every cycle of the conventional steam huff and puff and CO₂-assisted steam huff and puff.

It is worth noting that the method of injecting CO₂ gas first and high-temperature steam later was used in the implementation of CO₂-assisted steam huff and puff in this study. This method leverages two main mechanisms: firstly, the dissolution of CO₂ into heavy oil reduces its viscosity, thereby improving fluidity. Secondly, during the injection and early soak period, the expansion of the high-temperature steam chamber promotes the expansion and deeper penetration of co-injected CO₂. It is important to note that during the subsequent soak and production phases, steam condensation becomes the dominant process, releasing latent heat, which heats the reservoir. While this condensation leads to a pressure decline, the presence of gaseous CO₂ helps to mitigate this pressure drop, providing sustained drive energy and improving sweep efficiency.

The injection and production parameters of the model with bottom water were optimized after the optimization of the model without bottom water was completed. Obviously, the well layout should be different due to the presence of bottom water, so it needs to be optimized as well. Because of water channeling from the bottom, for the vertical well, several schemes with different perforation lengths were designed and the optimal scheme was selected by comparing the oil production results. As for the horizontal well, with the unchanging length of 100 m, the burial depth was adjusted: that is, the schemes of different water avoidance heights were set and selected by comparing the oil production results.

What calls for special attention is that, as

Table 4. Cyclic oil production (t) in each vertical well scheme.

Perforation Length	Cycle 1	Cycle 2	Cycle 3	Cycle 4	Cycle 5	Cycle 6
20 m	89	submerged	-	-	-	-
18 m	497	16	submerged	-	-	-
16 m	593	467	60	submerged	-	-
14 m	678	643	176	submerged	-	-
12 m	647	663	472	121	submerged	-
10 m	616	676	608	359	157	submerged

shows, the 14 m perforation length obtained the best oil recovery during the first huff and puff cycle, but soon, the oil production fell to zero, which meant that the well was submerged by water. Such a change cautioned us that the optimization method that based on the cyclic oil production results would be inaccurate. Thus, the optimization judgment was adjusted to the total oil production of six conventional huff and puff cycles. Schemes within which the wellbore was submerged in six cycles would be ignored. Optimization results are shown in Figure 5. The perforation length of the vertical well is 6 m from the top of the reservoir and the water-avoidance height of the horizontal well is 16 m: that is, the buried depth is 4 m from the top of the reservoir.

The original injection–production parameters are no longer applicable, obviously, when the well layout is changed. Therefore, the injection and production process parameters require optimization again for the condition with bottom water. The optimization method and steps are the same as without bottom water and results are shown in

Table 5. Injection and production parameters of reservoir with bottom water.

Parameters	Vertical Well	Horizontal Well
Steam injection intensity (t/m)	500	40
Steam injection rate (t/d)	500	1000
Soaking time (d)	10	15
Steam temperature (°C)	350	310
Steam quality	0.6	0.7
CO2 injection timing	Sixth cycle	Fifth cycle
CO2 injection intensity (m3/m)	8333	800
CO2 injection rate (m3/d)	20,000	5000
Liquid production rate (t/d)	20	20
Minimum bottom hole pressure of production well (kPa)	200	200

. Every injection and production parameter was unchanged in every cycle of conventional steam huff and puff and CO₂-assisted steam huff and puff.

The differences in injection and production parameters between two geological condition show that infinite bottom water generally made the steam and CO₂ injection intensity of the vertical well increase and the CO₂ injection intensity of the horizontal well decrease. Meanwhile, the soaking time decreased by about 70%. Furthermore, the duration of conventional steam huff and puff was significantly shortened. The duration of the horizontal well was shorter than the vertical well.

3.2. Analysis of Oil Production Performance

3.2.1. Influence of Bottom Water on Vertical Well

For the vertical well layout, six cycles of conventional steam huff and puff were performed for the no bottom water condition after the cold production stage, while only five cycles of conventional steam huff and puff were performed for the bottom water condition. Then, both conditions were performed for six cycles of CO₂-assisted steam huff and puff. The explanations of several key terms are shown in Appendix A. As can be seen from

Table 6. Development indicators of vertical well.

Development Indicators	Without Bottom Water	With Bottom Water
CO2 cumulative injection (t)	1601.99	553.98
Cumulative oil production during CO2-assisted huff and puff (t)	5140.42	846.40
Incremental oil recovery of CO2 (t)	2140.31	−14.46
Average EOR rate of CO2 (t/t)	1.27	−0.03
Cumulative water production during CO2-assisted huff and puff (t)	19,186.00	30,012.93
Incremental water recovery of CO2 (t)	7252.83	224.82
Average enhanced water recovery rate of CO2 (t/t)	4.31	0.13
Oil–water ratio during CO2-assisted huff and puff (t/t)	0.268	0.028

, the presence of bottom water will seriously reduce the oil production and greatly increase the water production, leading to an 89.6% reduction in the oil–water ratio compared to the no bottom water condition, eventually.

Incremental oil recovery of CO₂ is the oil recovery difference between conventional steam huff and puff and CO₂-assisted huff and puff, reflecting the EOR effect of injected CO₂. Figure 6 shows the cyclic incremental oil recoveries of CO₂ under two geological conditions. It is obvious that there are only three effective huff and puff cycles when bottom water exists (since the incremental oil recovery is negative after the fourth cycle), which is significantly shorter than the no bottom water condition. This indicates that when bottom water presents, the injected CO₂ not only has extremely poor EOR capabilities, but may even hinder the production of heavy oil, after multiple cycles of CO₂-assisted steam huff and puff. Despite the cyclic oil production of the no bottom water condition having a larger decline slope, it is still higher numerically than the bottom water condition in the same cycle.

On the other side, a higher incremental water recovery shows that CO₂ can improve water production when bottom water does not exist, but CO₂ plays an insignificant role in increasing water production when bottom water exists. This difference indicates that the bottom water is the main factor making the water production increase. These results demonstrate that the vertical well layout has a disappointing oil recovery effect while bottom water exists.

Taking the third cycle of CO₂-assisted steam huff and puff as an example, the oil production changes in a single cycle with and without bottom water are compared. As Figure 7 shows, during the initial stage of production of the no bottom water condition, the daily liquid production can reach 20 m³/d and the daily oil production keeps rising, but these two daily productions keep declining after a period of time. Interestingly, the daily liquid production under the condition of bottom water is almost continuously at full capacity over the whole production stage, but the daily oil production changes the same as no bottom water, rising at first and then decreasing, while the peak value is less than the no bottom water condition.

From Figure 8, we can see that the heavy oil in the lower part of the reservoir is difficult to use effectively because of the shorter perforation length of the bottom water condition. As a result, the oil production under the bottom water condition is more insufficient when the same number of huff and puff cycles are carried out. On the other hand, the presence of bottom water causes the water saturation around the well, generally 0.6–0.7, to greatly exceed the water saturation of no bottom water, which is less than 0.5. This indicates that the bottom water had flooded the wellbore as early as the third cycle of CO₂-assisted huff and puff, resulting in a sharp increase in water production and a decrease in the oil–water ratio, and weakening the stimulation effect of CO₂-assisted steam huff and puff, eventually.

A significant difference is provided by Figure 9: the formation pressure of no bottom water drops to about 200 kPa at the end stage of production during a single cycle. Moreover, pressure at the top reservoir is too low, making the formation water vaporize to form a gas chamber at the top. Such a phenomenon shows that the heat of the steam has been lost prematurely from the caprock, resulting in a decrease in steam heating efficiency. However, there is a stark difference: the formation pressure of the bottom water changes slightly at the end of a cycle compared with the initial stage, because of sufficient energy contained by infinite bottom water.

Combined with Figure 7, the production law of CO₂-assisted steam huff and puff can be obtained under two geological conditions. For the no bottom water condition, the formation pressure in the initial stage of production is supplemented by injected liquid and the production well can produce liquid at full capacity. However, the formation pressure drops with continuous production of formation fluid, resulting in a decrease in daily fluid production. As for the bottom water condition, the production well can produce liquid at full capacity throughout the whole cycle due to the high pressure, which is caused by sufficient energy from the bottom water. But the water cut of liquid production keeps rising as production proceeds, which makes the oil production decrease, eventually.

The temperature field variation in a single cycle is analyzed, revealing the influence of the bottom water on the heating effect of CO₂-assisted steam huff and puff through Figure 10. As can be seen from the graph, the residual temperature around the well with the bottom water condition is higher than the no bottom water condition at the end of steam injection, because the former’s steam injection strength is higher than the latter. However, the steam heating radius of the bottom water condition is about 20 m lower in the horizontal direction than the no bottom water condition. As for the vertical direction, the steam under the bottom water condition is difficult to spread to the heavy oil below, due to the short length of the perforation. Thus, the heating efficiency is ultimately lower than in the no bottom water condition. In addition, it is easy to observe that the upward intrusion of the bottom water cone continuously decreases the temperature in the heating radius as production continues, which makes the viscosity of the heavy oil there increase. This significant outcome indicates that the expansion of the steam chamber would be suppressed by the bottom water and the heating range would be continuously reduced as the water encroaches. Although the residual temperature around the wellbore is higher than the no bottom water condition, the heating efficiency is still not as good as the latter, leading to a reduction in the heavy oil’s production capability, ultimately.

3.2.2. Influence of Bottom Water on Horizontal Well

As for the horizontal well layout, eight cycles of conventional steam huff and puff were performed for the no bottom water condition after the cold production stage, while four cycles of conventional steam huff and puff were performed for the bottom water condition. Then, six cycles of CO₂-assisted steam huff and puff are performed for both conditions. Both development indicators are shown in

Table 7. Development indicators of horizontal well.

Development Indicators	Without Bottom Water	With Bottom Water
CO2 cumulative injection (t)	1601.97	897.60
Cumulative oil production during CO2-assisted huff and puff (t)	15,184.43	3287.08
Incremental oil recovery of CO2 (t)	6880.43	1512.56
Average EOR rate of CO2 (t/t)	4.09	0.90
Cumulative water production during CO2-assisted huff and puff(t)	26,498.50	51,754.46
Incremental water recovery of CO2 (t)	10,149.20	8388.15
Average enhanced water recovery rate of CO2 (t/t)	6.03	9.35
Oil–water ratio during CO2-assisted huff and puff (t/t)	0.573	0.064

. The explanations of several key terms are shown in Appendix A. It can be seen that, similarly to the vertical well layout, bottom water will seriously reduce the oil production and greatly increase the water production. The oil–water ratio, whose variation is similar to the vertical well layout as well, decreased by 88.8% compared with the no bottom water condition. But, on the whole, for the bottom water condition, oilfield development indicators of the horizontal well layout generally get greater numbers than that of the vertical well layout.

According to Figure 11, the no bottom water condition has only five effective huff and puff cycles, since the incremental oil recovery is negative after the sixth cycle. Meanwhile, its oil recovery decline slope is larger than the vertical well. But, on the other side, effective huff and puff cycles with the bottom water condition, unlike the vertical well, can last longer with better oil recovery efficiency. What interests us is that the incremental oil recovery of both formation conditions account for 45% of the cumulative oil production. Such a value indicates that the EOR effect of CO₂ will not be influenced by bottom water.

However, it is important to notice that the enhanced water recovery rate of CO₂ under the bottom water condition is 1.5 times as much as no bottom water condition, and the gap is even bigger compared to the vertical well. This peculiar difference reveals another characteristic of CO₂: that it can enhance not only oil recovery, but also water recovery. As a result, for the bottom water condition, the oil–water ratio of the horizontal well is very low, but still higher than the vertical well. Thus, the horizontal well layout is more suitable in heavy oil production if bottom water exists.

When produced with the horizontal well, the oil above the well will be produced first, due to gravity. Thus, Figure 12 presents that the no bottom water condition has a bigger area of low remaining oil saturation above the horizontal well, while the horizontal well in the bottom water condition is forced to rise. On the other side, the remaining oil saturation in the area under the horizontal well is generally about 0.4 when bottom water exists, while the remaining oil saturation of the no bottom water condition in that area is above 0.6.

Combined with the analysis of the water saturation distribution field, the increase in water saturation is small under the no bottom water condition, while the increase is great if bottom water presents, which is generally above 0.55, indicating that the water crest has already penetrated into the wellbore. This remarkable sign may explain why the remaining oil saturation under the horizontal well decreases significantly under the condition of bottom water. The heavy oil under the horizontal well is pushed into the wellbore by the bottom water’s sufficient energy during the early stage of production. But, the heavy oil-producing degree in this area still cannot make up for the decline in the oil–water ratio and the decrease in stimulation efficiency, which caused by huge amount of water production. Compared with the vertical well under a corresponding formation situation, it is clear that the horizontal well has greater conformance efficiency, which means that it has better oil production than the corresponding vertical well scheme.

From the change in pressure field shown in Figure 13, it is apparent that the formation pressure around the wellbore at the end of a cycle has dropped to about 200 kPa when there is no bottom water, which is consistent with the case of the vertical well. Moreover, the existence of the gas chamber at the top has also occurred. As for the case of the bottom water, the formation pressure at the end of a cycle is still huge because the infinite bottom water contains sufficient energy, which has few differences from the initial state, even after multiple cycles of huff and puff.

As can be seen from Figure 14, steam overlap occurs at the end of steam injection stage of no bottom water condition, due to gravitational differentiation, which leads to a tendency of the temperature field to spread to the upper part of the horizontal well, where the heating temperature is generally around 200 °C and keeps decreasing with the progress of production continuing. As for the bottom water condition, the temperature is about 350 °C around the wellbore after steam injection, even higher than injected steam’s temperature, which is 310 °C. That is because there is a phenomenon of pressure build-up occurring—it is hard for the steam to flow and gather here, making the pressure and temperature increase.

It is worth noticing that the temperature above the well of the bottom water condition after 4 months is about 270 °C, while the temperature below the well is only 180 °C. The latter decreases significantly than the former. Combined with the migration law of the bottom water and water saturation distribution, it is easy to infer that a large amount of low temperature fluid migrates here and causes the temperature to drop rapidly due to the bottom water upwelling. Meanwhile, further analysis shows that the pressure produced by the bottom water upwelling inhibits the downward diffusion of the steam chamber to some extent by observing the variation in pressure field together. Moreover, the upwelling can compress the steam chamber as production continues.

3.3. Analysis of CO₂ Storage Capability

As a CCUS measure, CO₂-assisted steam huff and puff can store CO₂ efficiently and reduce carbon emissions.

Table 8. CO₂ storage capability of each scheme.

	Vertical Well		Horizontal Well
	No Bottom Water	With Bottom Water	No Bottom Water	With Bottom Water
Total CO2 injection (t)	1602.0	534.0	1602.0	897.6
Total CO2 production (t)	1548.8	220.2	1517.8	33.8
CO2 storage (t)	53.2	313.8	84.2	863.8
CO2 storage efficiency	3.3%	58.8%	5.3%	96.2%

shows the CO₂ storage capability of four schemes. The explanation of the CO₂ storage efficiency is shown in Appendix A. It is remarkable that CO₂ storage under the bottom water condition increases significantly, compared with the no bottom water condition. The CO₂ storage efficiency of the horizontal well is about 63% larger than the vertical well. These interesting results indicate that the horizontal well has a better CO₂ storage ability in CO₂-assisted steam huff and puff when bottom water exists.

Why is the CO₂ storage efficiency of the no bottom water condition so low? Further analysis of the formation temperature, pressure and CO₂ distribution in the reservoir shows the answer. Take the first cycle of CO₂-assisted steam huff and puff in the horizontal well as an example. From Figure 15, we can see that there is a large amount of CO₂ existing in the formation in the bottom water condition, while another condition has little CO₂ during the oil production period. According to our analysis of pressure distribution above, we know that the bottom water is able to maintain the formation pressure, even making it beyond the initial pressure. Hence, the formation pressure in the whole production process is higher than the critical pressure of CO₂—7380 kPa. Also, the formation temperature is much higher than the critical temperature of CO₂—31.04 °C. That is, CO₂ in the bottom water formation is supercritical, and it has better mobility than gaseous CO₂ at a high pressure level in the reservoir and other functions that greatly enhance oil recovery, with its reduced viscosity, increased diffusivity, and liquid-like density [59,60]. On the other hand, the formation without bottom water could not maintain a high pressure. The CO₂ inside is gaseous and easy to produce before it fully dissolves with the crude oil, which is unable to proceed with effective CO₂ storage.

Above all, due to the high temperature brought by the steam and the high pressure maintained by the bottom water, although gaseous CO₂ at room temperature is injected, CO₂ will be supercritical under the bottom water condition and its storage efficiency is much larger than that of the no bottom water condition. This analysis indicates that CO₂-assisted steam huff and puff technology has an effective CO₂ storage capability while bottom water exists.

4. Sensitivity Analysis

In the process of reservoir development, porosity, permeability and oil saturation, as the core parameters characterizing the reservoir’s physical properties and fluid distribution, will significantly affect the prediction accuracy of oil production. This paper, by designing a multi-factor sensitivity numerical simulation, quantitatively analyzed the influence degrees and interaction laws of these three key parameters on the oil production and the incremental oil recovery of CO₂ and CO₂ storage efficiency during the process of bottom water heavy oil exploitation, using CO₂-assisted steam huff and puff technology in order to determine the main control factor at different development stages and provide a scientific basis for the site selection and adjustment of bottom water reservoir exploitation.

For the formation porosity, five schemes of 0.1, 0.2, 0.3, 0.4 and 0.5 were designed. For the formation permeability, five schemes of 500 mD, 1000 mD, 2000 mD, 3000 mD and 4000 mD were designed. For the initial oil saturation, six schemes of 0.5, 0.6, 0.7, 0.8, 0.9 and 1.0 were designed to numerically simulate the reservoir models of heavy oil with bottom water in the horizontal well and vertical well, respectively. The injection and production parameters of the two well layouts were consistent with those mentioned above, with no other changes. To unify the huff and puff effect, after the cold production stage, each scheme uniformly conducted two cycles of conventional steam huff and puff and two cycles of CO₂-assisted steam huff and puff. After the simulation of each scheme was completed, the influences of porosity, permeability and oil saturation on cumulative oil production, incremental oil recovery of CO₂ and CO₂ storage efficiency were compared.

4.1. Analysis of Porosity

Figure 16 shows the influence of porosity changes on cumulative oil production, incremental oil recovery of CO₂ and CO₂ storage efficiency. For the vertical well, as porosity increases, the pore volume in the formation increases too, which means the oil storage volume increases as well. Therefore, the oil production effect is better. However, the incremental oil recovery of CO₂ decreases with an increase in porosity, due to the latter leading to an increase in the oil storage in the formation while the total injection of CO₂ remains unchanged. Therefore, the unit oil volume dissolves a smaller amount of CO₂, resulting in a decline in the oil-enhancing effect of CO₂. The incremental oil recovery of CO₂ is lower than zero when the formation porosity is higher than 0.4, indicating that CO₂ no longer has the oil-increasing effect in this situation. This is because, on the one hand, the high porosity enables CO₂ to flow more widely in the formation pores. Coupled with the high pressure gradient caused by the bottom water, CO₂ quickly passes through the oil layer to reach the top of the reservoir, thereby reducing the contact time with the heavy oil. On the other hand, the high porosity significantly increases the recovery rate of conventional thermal recovery, indirectly reducing the oil-enhancing capacity of CO₂. The CO₂ storage efficiency continuously decreases with the increase in porosity. The CO₂ storage efficiency in the 0.5 porosity scheme decreases by 12% compared with that in the 0.2 porosity scheme. This indicates that the higher the porosity, the more difficult it is for CO₂ to dissolve with oil and remain in the formation; thus, the easier it is to produce. This is consistent with the above conclusion that the oil-enhancing effect of CO₂ decreases.

For the horizontal well, the variation in the incremental oil recovery of CO₂ is different. It reaches the highest when the porosity is 0.2, then shows a fluctuating trend with the increase in porosity. The variation trend of the CO₂ storage efficiency is also similar to the incremental oil recovery, reaching the peak at 0.2. However, the incremental oil recovery of CO₂ accounts for 14% of the cumulative oil production in the 0.2 porosity scheme, while it only accounts for 6% of the cumulative oil production in the 0.5 porosity scheme. But, the incremental oil recovery and enhanced oil recovery rate of CO₂ in the two schemes, considering their definitions, do not have much difference. This might imply that for the horizontal well, the injection volume of CO₂ needs to be increased to enhance the oil-enhancing effect when the porosity is greater than 0.3.

4.2. Analysis of Permeability

Figure 17 shows the impact of the permeability changes. For the vertical well, the impact of the permeability changes on cumulative oil production is not remarkable, but it significantly affects the incremental oil recovery and storage efficiency of CO₂. The excessively low permeability leads to a decrease in the conformance efficiency of the injected CO₂ gas, making it difficult to dissolve in the heavy oil around the wellbore and also difficult to store in the formation. As a result, it is produced in large quantities during the production stage, making the incremental oil recovery and storage efficiency low. With the increase in permeability, the CO₂ storage efficiency keeps rising. However, the incremental oil recovery reaches its peak in the 2000 mD scheme and then continuously decreases. When the formation permeability reaches 4000 mD, the incremental oil recovery of CO₂ is less than 0, indicating that the oil-enhancing effect of CO₂ disappears at this time. The high permeability leads to high fluidity of CO₂, which can rapidly diffuse along the seepage channel to the distant part of the wellbore and dissolve with the heavy oil there. On the one hand, less CO₂ is dissolved in the heavy oil near the wellbore, reducing the oil-enhancing effect of CO₂. On the other hand, the heavy oil in the area far from the wellbore dissolves CO₂, but fails to transfer to the wellbore and be produced. Therefore, a large amount of CO₂ can be stored in the reservoir.

For the horizontal well, the influence on the CO₂ storage efficiency is similar to that of the vertical well: relatively low in low-permeability reservoirs and increasing with permeability increasing. The CO₂ injected into the horizontal well has a wider-spread range. Meanwhile, it can alleviate the pressure shock caused by the bottom water cresting and greatly improve the solubility of CO₂ near the wellbore. Furthermore, the production efficiency of the horizontal well is significantly higher than that of the vertical well, with a wider range of utilization for heavy oil. Therefore, CO₂ maintains a good oil-enhancing effect in all permeability schemes, reaching a peak at 3000 mD, and then slightly decreases with the increase in permeability.

4.3. Analysis of Initial Oil Saturation

The influences of the change in initial oil saturation are shown in Figure 18. The initial oil saturation of the vertical well shows an evident positive correlation with the cumulative oil production and the incremental oil recovery of CO₂. The CO₂ storage efficiency reaches the peak when the initial oil saturation is 0.6 and then decreases significantly, because at a high saturation, more CO₂ dissolves in the oil around the wellbore, improving the oil-enhancing effect and incremental oil recovery of CO₂. On the other hand, it also makes CO₂ easier to produce, leading to a low CO₂ storage efficiency. Thus, when the initial oil saturation is 1.0, the CO₂ storage efficiency decreases by 21% compared with that while the initial oil saturation is 0.6.

The variation trend of incremental oil recovery of CO₂ has distinct differences while using the horizontal well. It does not continuously increase with the initial oil saturation rise, but it fluctuates constantly when the saturation is 0.6–0.8 and then slowly decreases again at a higher saturation. CO₂ storage efficiency, similar to that of the vertical well, is higher at a low saturation and slightly decreases with the increase in saturation. This might be because the conventional steam huff and puff of the horizontal well has a larger swept range and better oil production effect under high saturation. Therefore, it still indirectly leads to a decrease in the oil-enhancing effect of CO₂, although more CO₂ dissolves in the oil around the wellbore.

4.4. Influence Levels of Sensitive Parameters

In order to clarify the influence levels of each sensitive parameter on the oilfield development indicators,

Table 9. Sample variances of each development indicator.

	Vertical Well			Horizontal Well
	Cumulative Oil Production	Incremental Oil Recovery of CO2	CO2 Storage Efficiency	Cumulative Oil Production	Incremental Oil Recovery of CO2	CO2 Storage Efficiency
Porosity	1,705,404	2529	0.0036	11,178,992	22,780	0.0022
Permeability	4866	946	0.0112	6806	5670	0.0186
Initial oil saturation	3,756,104	16,994	0.0044	32,945,884	1752	0.0004

lists the sample variances of the oilfield development indicators under each type of sensitive parameter scheme. For the vertical well, oil saturation has the greatest influence on the cumulative oil production and incremental oil recovery of CO₂, followed by porosity, and permeability has the smallest influence. However, permeability has the greatest influence on the CO₂ storage efficiency, while the porosity and oil saturation have relatively small influences. For the horizontal well, oil saturation has the greatest influence on the cumulative oil production, porosity has the greatest influence on the incremental oil recovery of CO₂, and permeability has the greatest influence on the CO₂ storage efficiency.

Overall, the oil saturation has a significant impact on the overall oilfield development indicators in heavy oil production projects with bottom water. The permeability has a significant impact on the CO₂ storage efficiency. The sensitive parameters which affect the incremental oil recovery of CO₂ are related to the well layouts.

5. Conclusion and Discussion

5.1. Conclusions

In this paper, we built typical offshore heavy oil geological models with and without bottom water to study the oil and water flow process, CO₂ storage capability and maximal effect of infinite bottom water during CO₂-assisted steam huff and puff. In particular, the oil recovery analysis, parameter optimization and CO₂ storage efficiency were conducted for the condition, using vertical and horizontal well layouts. The following conclusions were drawn from these simulation results and analysis.

(1)

The formation pressure will not change significantly and the steam chamber radius can be reduced by about 50% when bottom water exists. Compared with the no bottom water condition, infinite bottom water causes the length of the perforation section of the vertical well to be shortened by 70%, and it raises the horizontal well to 1/5 of the thickness from the top of the reservoir.

(2)

Bottom water would make the soaking time of both the vertical and horizontal wells reduce by 70% at most. But the CO₂ injection intensity and rate of the vertical well need to be increased and those in the horizontal well need to be reduced. Meanwhile, the horizontal well layout would have fewer effective conventional steam huff and puff cycles than the vertical well layout.

(3)

The oilfield development indicators, effective production duration and the EOR effect of CO₂ of the horizontal well are generally higher than that of the vertical well, with each indicator increasing by 40–80%. The CO₂ storage efficiency in the horizontal well is 63% larger than in the vertical well. Thus, horizontal wells are more suitable for the production of heavy oil reservoirs with bottom water by CO₂-assisted steam huff and puff.

(4)

Analysis of bottom water formation sensitivity parameters shows the advantageous formation conditions under each well layout. The oil saturation has a significant impact on the overall oilfield development indicators. The permeability has a significant impact on the CO₂ storage efficiency. The sensitive parameters which affect the incremental oil recovery of CO₂ are related to the well layouts.

5.2. Discussions

This study primarily employs a conceptual model with an infinite aquifer to elucidate the fundamental mechanisms of CO₂-assisted steam huff and puff in bottom water reservoirs. While this approach provides clear insights into the extreme influence of a strong aquifer, it also opens avenues for further discussion of the key practical considerations.

(1)

Implications of water volumetric multiple of aquifer

The use of an infinite aquifer model in this work represents a boundary case of maximum aquifer support. In practical field applications, the aquifer size is finite, characterized by a water volumetric multiple. The magnitude of this multiple would significantly influence the reservoir pressure maintenance and the severity of water coning. A weaker aquifer (lower water multiple) would likely lead to a more rapid pressure decline, potentially reducing water production but also shortening the effective production period of the well. Conversely, a stronger aquifer would align more closely with our findings, providing better pressure support but posing greater challenges in controlling the water cut. Future research incorporating aquifers with varying strengths is essential to establish a quantitative relationship between water volumetric multiples and development indicators, thereby optimizing the development strategies for specific reservoir conditions.

(2)

Long-term stability of CO₂ geological sequestration

This study focuses on the near-term retention of CO₂ that remains in the reservoir following the enhanced oil recovery process, rather than its long-term geological sequestration. Long-term sequestration involves complex factors extending beyond our present scope, such as mineral trapping and caprock integrity over centuries. Consequently, our analysis is confined to the CO₂ that is not produced during the production phase.

Under bottom water conditions, the high-pressure environment helps to maintain CO₂ in a dense supercritical state, promoting its retention. Nevertheless, the long-term security of the retained CO₂ requires further investigation. Potential risks include the integrity of the caprock and sealing faults under the cyclic thermal and pressure stresses induced by the huff-and-puff process. A comprehensive assessment combining long-term predictive modeling with site-specific geological data is recommended for future work to fully evaluate the containment reliability.