The Influence of CO₂ Injection Timing and Well Placement on Storage Indicators in a Producing Natural Gas Reservoir with High CO₂ Content

Abstract

Carbon capture and storage is recognized as the best method for CO₂ emission mitigation in cases where the emissions cannot be avoided by switching to renewable energy or by increasing efficiency. In the Republic of Croatia, there are a number of gas-condensate reservoirs with a high content of CO₂ that are adequate candidates for CO₂ storage. The injection starts at selected different points before the end of production and might bring financial benefits as a result of CO₂ injection patterns and efficient permanent storage. Employing a model of a typical reservoir, various scenarios of CO₂ injection are examined to assess the impact of injection well positioning and timing on the reservoir’s sustainability for permanent CO₂ storage. The main indicators used for the assessment are additional hydrocarbon recovery, retention, and storability.

1. Introduction

In recent decades, the European Union had set ambitious plans to reduce CO₂ emissions and other greenhouse gases into the atmosphere. Although an increase in the share of renewable energy sources and the reduction in fossil fuel consumption can contribute toward these goals, it is unreasonable to believe that any significant emission reduction would be achieved without the implementation of carbon capture and storage (CCS) and carbon capture, utilization and storage (CCUS) technologies [1,2]. Reduced use of fossil fuels in a cost-effective way creates an opportunity for the development of CCUS-related technologies. Those technologies have the potential to simultaneously reconcile energy security and ecology [3]. The petroleum industry is traditionally seen as a big polluter, contributing 8% to direct global emissions and 33% to indirect global emissions [4]. This is the main reason why new approaches and innovative solutions are explored within the petroleum industry: in order to reduce its significant environmental impact.

The study presented in this paper (CO₂ storage in gas reservoirs) supports sustainable energy practices by repurposing the existing infrastructure, minimizing new hydrocarbon exploration needs, and achieving gate-to-gate carbon negativity, where stored CO₂ exceeds CO₂ produced over the reservoir lifecycle. In this way, the proposed concept delivers reduced environmental impacts through GHG reduction, economic feasibility via EGR cost savings, and social benefits by supporting fossil-to-renewable transitions without disrupting energy supply.

CCS technology represents the geological storage of CO₂, which aims to capture CO₂ from different industries, transport it, and retain it underground permanently, to prevent its release into the atmosphere. There are numerous studies examining the suitability of depleted oil and gas reservoirs and large and deep saline aquifers for CO₂ storage [5,6,7]. At the same time, hydrocarbon reservoirs are considered to be more reliable for geological storage, due to their proven ability to retain fluids for a long time. Far more data about subsurface properties, and therefore about CO₂ storage suitability, is acquired through producing oil and gas wells, compared to deep saline aquifers which are characterized by a more general approach, and eventually, from the data from exploration wells that were not drilled into in an oil or gas reservoir. Due to growing awareness of CCS as a perspective carrier for achieving the ambitious goals established by the Paris Agreement [8] and the high price of CO₂, which became less volatile after the Green Deal, positive changes can be expected in the coming period [9]. One of the motives was to examine early CO₂ injection, not only to store greenhouse gas, but also to improve the efficiency of the reservoir, including improving CO₂ storage aspects in a sustainable manner and reducing the need for new hydrocarbon exploration.

The concepts of chemical utilization imply the conversion of carbon dioxide into another chemical compound, while the concepts of physical utilization with simultaneous storage mostly refer to methods for increasing the hydrocarbon production, whereby CO₂ is stored in underground formations without chemical structure alteration [10]. Utilization is practically the only way to offset high CO₂ capturing and storage costs [11], and the benefit of trapping significant amounts of CO₂ makes such simultaneous CO₂ storage and (to some extent) enhanced hydrocarbon recovery attractive from both an emission reduction and a financial point of view [12]. Many years of successful CO₂-enhanced oil recovery (CO₂-EOR) practice around the world, including two projects for enhancing oil recovery (EOR) in Croatia, is a guarantee of the advanced level of development of these technologies, meaning their technological readiness level (TRL) is at its maximum, considering the Global CCS Institute’s classification [13]. Furthermore, economic profitability and retention of large amounts of CO₂ underground during CO₂-enhanced hydrocarbon production make these options extremely attractive. Unlike CO₂ injection for enhanced oil recovery, it is hardly possible to enhance the recovery of natural gas reservoirs because of the CO₂ and methane mixing, leading to the rapid breakthrough of CO₂ [14,15]. These challenges can be mitigated by finding the optimal position of the injectors and producers—it is advisable to place the injection wells deeper in the reservoir, while production wells should be perforated in the upper layers to exploit the gravity segregation effects [16]. This position management logic is based on CO₂ PVT (pressure–volume–temperature) behavior. CO₂ density can be two to six times higher than natural gas density in reservoir conditions, and at the same time, CO₂ has lower mobility compared to natural gas (methane). These properties cause CO₂ to migrate towards lower (deeper) parts of the reservoir. Additionally, the difference in these properties should enable a slower exchange of CO₂ and methane. Furthermore, CO₂ has a higher solubility factor in brine than methane (considering reservoir conditions), which affects the CO₂ breakthrough, i.e., it delays it [17,18]. It was found [19] that besides the careful positioning of the injection and production wells, another crucial parameter for enhanced gas recovery (EGR) design is the decision on the injection rate, which is highly dependent on the reservoir’s anisotropy (i.e., the difference between horizontal and vertical permeabilities). Furthermore, CO₂–methane mixing could be hindered by the injection of carbonated water slugs [20].

All these previously published findings confirm that, for enhanced natural gas production, the favorable effects such as oil-swelling and oil viscosity reduction are not appearing in the case of CO₂ injection into gas reservoirs. However, by injecting CO₂ into a (mature) producing gas reservoir, we expect slower pressure depletion compared to gas production without CO₂ injection, and also better conditions for CO₂ injection, because gas production wells act as “pressure relief” wells. Consequently, despite the fact that significant total gas production will not increase, we focused on the assumption that by pressurizing the reservoir at the injection well, about the same amount of total gas production can be achieved earlier (than in the case without injection). This concept entails a new approach to CO₂ storage, advancing sustainable greenhouse gas mitigation in producing reservoirs and balancing efficient hydrocarbon recovery, including an early start of emission reductions with long-term environmental and economic benefits.

The issues of injecting CO₂ into gas reservoirs are still insufficiently investigated and leave a lot of room for research, especially when considering reservoirs with high CO₂ content that are already producing. This study used a natural gas reservoir with a high concentration of CO₂ in its original composition to perform a history-matching of production data, hence enhancing the accuracy of CO₂ injection simulations. Given the scarcity of the existing literature on the subject, this study primarily focuses on the CO₂ storage rather than the gas-production enhancement. However, the underlying premise is that initiating CO₂ injection during gas production would yield greater (cost) effectiveness. This can be explained by early installation of CO₂ injection equipment, data gathering, and a relatively higher depletion of the reservoir, which would result in additional capacity being available for CO₂ storage.

While CO₂ injection in a dry gas reservoir can increase the recovery by 4–12%, and it is advisable to start EGR only after the primary production ceases, additional recovery from a wet gas reservoir can be as high as 40% if the injection is started early enough to avoid condensate dropout (related to the pressure drop) in the reservoir [21]. The main challenge of CO₂ injection (both for EGR and ‘pure’ storage) is low reservoir pressures that are typical for depleted gas reservoirs, because at these low pressures, Joule–Thomson cooling effect could take place, leading to further operational problems, such as hydrate-forming and reduced well injectivity: i.e., an increased wellbore skin factor [22]. Considering that the reservoir fluid examined in this study is a gas-condensate reservoir (which is condensing because of high CO₂ content), the additional recovery rate of 40%, found in the literature, was the main reason why EGR scenarios were considered in addition to assessing the storage efficiency. Furthermore, the mentioned issues that are related to low reservoir pressures are the main motivation for examining the influence of different injection timings.

2. Materials and Methods

The goal of this work was primarily to observe the impact of the injection well position and (since such problem statement has not been found in the literature) the start time of the CO₂ injection on the reservoir’s suitability for CO₂ long-term storage, but the impact on the hydrocarbon recovery was also observed. As reservoir behavior depends on the injected and produced fluid characteristics: a short overview was given in that regard.

2.1. Characteristics of CO₂ Injection into Gas Reservoirs

Due to significantly different CO₂ and natural gas properties in reservoir conditions, they usually do not mix easily, which reduces the likelihood of natural gas contamination. Gravitational segregation causes the CO₂ (due to its density) to sink towards the reservoir bottom. This way, a “gas cushion” is formed that can enhance the production of natural gas through piston-like displacement (Figure 1): i.e., CO₂ “pushes” natural gas towards production wells [23]. Furthermore, CO₂ is more viscous than natural gas in the reservoir conditions, leading to a favorable mobility ratio and an efficient displacement process [24].

Figure 1. Displacement mechanism scheme for CO₂ and natural gas [3].

Different PVT (pressure–volume–temperature) behavior of CO₂ and natural gas suggests that CO₂ could have several “roles” in reservoir processes. It contributes to the repressurization of the reservoir (i.e., it helps to restore the original reservoir pressure) and displaces natural gas at the same time. Earlier simulation studies ([25,26]) indicate that an additional 5 to 15% of natural gas can be produced through CO₂-enhanced gas recovery (EGR). The results of the K12-B pilot project have shown that an additional 0.03 to 0.05 t of gas can be recovered per one ton of injected CO₂ [27], leaving room for storing CO₂, even while the gas reservoir is still in production. This is in line with the understanding that depleted gas reservoirs have greater CO₂ storage capacity in comparison with depleted oil reservoirs, as the recovery factor (which is directly proportional to the reservoir’s CO₂ storage capacity) in gas reservoirs can be around 60%, while the ultimate oil primary recovery rarely exceeds 30%. Volumetric equations can be used in the initial estimation of the reservoir’s CO₂ storage capacity with a satisfactory accuracy [28], especially in dry gas reservoirs [29]. It was found that the CO₂ injection rate does not affect the overall storage capacity of a reservoir [30].

2.2. The Research Overview

A geological model of a producing reservoir was built with Schlumberger Petrel software (version 2024.4), which is also used to run compositional Eclipse (E300) implicit simulations [31]. Among commonly used cubic equations of state, such as Peng–Robinson [32], Redlich–Kwong [33], or Stryjek–Vera [34], in this research, the PR equation of state (EoS) was selected for describing the fluid behavior, based on the available experimental data obtained within the previous research [35]. Except for the selected EoS, estimation of the dynamic injectivity changes depends on the additional details related to reservoir properties and the CO₂ injection strategy [36]. The injectivity primarily depends on reservoir permeability, effective layer thickness, heterogeneity, and reservoir and injection pressure. Relative permeability for CO₂ increases in regions with increased CO₂ saturation ([14,37]), leading to better injectivity. Low injectivity can cause an increase in pressure, meaning that more energy for CO₂ compression is needed, which directly increases the injection costs [38]. Due to differences in relative permeability, supercritical CO₂ can displace natural gas, making the process immiscible [38]. As described in the reviewed literature ([39,40]), many challenges arise when attempting to incorporate detailed microscopic effects (e.g., wettability, miscibility, and interfacial tension changes between rock, natural gas, brine, and CO₂) into a full-scale (reservoir) simulation model.

Different development strategies (dynamic models) were defined, regarding the injection timing and injection well placement.

The research was conducted through eight steps:

• Develop a reliable simulation model by achieving satisfactory field pressure and production history matching.
• Create prediction strategies for base cases (business as usual, BAU) where the current production strategy continues (no CO₂ injection) until the production’s end, when the reservoir is repurposed as CO₂ storage.
• Create predictive strategies that include simultaneous CO₂ injection and gas production.
• Assess the impact of the CO₂ injection timing and the injection well selection on the storage capacity.
• Evaluate different CO₂ injection strategies in terms of permanently stored amount of CO₂.
• Compare the defined injection strategies with the BAU scenarios, regarding the hydrocarbon recovery.
• Determine the impact of the CO₂ injection timing and the injection well selection on the CO₂ breakthrough and reservoir fluid composition changes.
• Assess the effect of CO₂ injection in BAU cases (‘pure’ storage after reservoir depletion) on the storage potential, including reservoir pressure.

Prior to describing the details of the simulation scenarios, a short rationale is given to put the examined reservoir into a broader context, and some theoretical background is introduced for better understanding of the results and their meaningful interpretation.

2.3. Theoretical Background

Although the CO₂ storage history started in the early 1970s [41], most of the development projects have not become operational until today. However, there are some indications of the increasing popularity of CCS, mostly in North America and Europe [42]. Depleted oil and gas reservoirs are currently considered to be the best option for CO₂ storage sites from an infrastructural, reservoir reliability, and economic point of view, although capacity-wise, aquifers seem more advantageous [43].

With the rising importance of a natural gas supply on one side, and emerging stricter legislation on GHG emissions on the other, special cases of natural gas reservoirs with high initial CO₂ content might become a controversial issue, since it is virtually impossible for such reservoirs to be exploited in a carbon neutral way without CO₂ injection.

The reservoir from this research has around 50% mol of CO₂ in the initial reservoir fluid (gas-condensate) composition, which makes it an interesting example for storage capacity assessment (expressed through two indicators—retention and storability), and is especially interesting from the hydrocarbon recovery point of view, considering that CO₂ is not an efficient fluid for natural gas production enhancement, as already explained in the Introduction.

As this research is not focused on the typical EGR process, the usual indicators, such as additional gas recovery, are not relevant, and the efficiency of CO₂ permanent storage for each scenario is evaluated through relevant indicators, such as retention (Equation (1)) and storability (Equation (2)): parameters that account for CO₂ that is cumulatively injected and produced (considering the period of simultaneous production and injection). Retention represents the difference between the injected and produced CO₂, and storability is expressed as the ratio of this difference (i.e., retention) and produced CO₂ [44]. The primary focus of this research is permanent CO₂ storage, while hydrocarbon recovery enhancement was considered only as a secondary indicator. However, both parameters should have positive values, and their high values are a good indicator that carbon negativity can be achieved in the scenarios (more CO₂ injected than produced). To fully assess the carbon negativity of the overall process, a thorough life cycle assessment should be conducted, which is out of the scope of this research, so for the purpose of this work, carbon negativity refers solely to the gate-to-gate comparison of CO₂ that is produced and permanently stored.

Retention = injected CO₂ volume–produced CO₂ volume in the injection period, m³.

(1)

Storability = Retention/Produced CO₂ volume in the injection period, (-)

(2)

While retention can be directly interpreted as the amount of the injected CO₂ that is permanently stored (retained) in the reservoir, this indicator is not sufficient to assess the storage efficiency in the simultaneous injection and production period, especially for reservoirs with high initial CO₂ content. Storability gives a relative relation between the stored and produced CO₂ amount: i.e., it quantifies the storage efficiency in terms of the CO₂ being permanently stored that is expressed as a percentage of the CO₂ produced, eliminating the ambiguity that might arise from expressing only the absolute value of the CO₂ retained. Although the permanently stored amount of CO₂ could be observed in relation to the injected CO₂, for the purpose of this research, it is more useful to observe it in the context of the amount of CO₂ produced.

2.4. The Research Workflow and the Reservoir Model Details

The model developed in the Schlumberger Petrel simulation software (version 2024.4) was initialized using the available geological model of a real reservoir, including known reservoir and fluid data (Table 1).

Table 1. Initial reservoir fluid mole composition.

Component i	zi, %
Nitrogen (N2)	1.98
Carbon Dioxide (CO2)	48.95
Methane (C1)	45.41
Ethane (C2)	1.66
Propane (C3)	0.36
iso-Butane (i-C4)	0.13
n-Butane (n-C4)	0.22
Iso-Pentane (i-C5)	0.07
n-Pentane (n-C5)	0.07
Hexane (C6)	0.16
Heptane (C7)	0.15
Octane (C8)	0.11
Nonane (C9)	0.10
Higher hydrocarbons (C10+)	0.63

For those properties that were not sufficiently tested, appropriate correlations were applied. Production strategies were created by choosing different well production constraints. Figure 2 conceptually shows the workflow in this research, with green and orange boxes containing input data that is the same for all scenarios (strategies) and gray boxes describing the differences between the scenarios. The obtained results were compared in order to establish the effect of different production and injection well selections and different injection timings on the reservoir response. The main parameters of interest observed include the amount of CO₂ injected and produced (the difference between the two being the permanently retained CO₂), hydrocarbon production, and the reservoir pressure.

Figure 2. Simulation workflow.

The used model has a corner-point grid of NX × NY × NZ = 28 × 30 × 9 = 7560 grid blocks to describe the structure with dimensions DX × DY × DZ = 129.1 × 124.5 × 26.1 m.

Figure 3 displays the wells utilized in the model, along with the porosity model derived from actual field data (namely well logging and seismic).

Figure 3. Reservoir porosity distribution.

Reservoir permeability in horizontal directions was modeled with a log-normal distribution, using a mean value of 5 mD with a standard deviation of 1 mD, while vertical permeability was also described with a log-normal distribution, but with the mean and standard deviation being 10 times smaller. Correlations were used for rock compaction [45], relative permeability (using Corey exponents that are typical for natural gas reservoirs, Figure 4), and brine properties (based on water salinity of 30,000 ppm).

Figure 4. Relative permeability curves used in the model.

The gas–water contact was determined based on well-logs, and it was set at an absolute depth of −2525 m (Table 2). Reservoir temperature and pressure at the contact depth are also given in Table 2.

Table 2. General reservoir data.

Gas–water contact (reference depth)	−2525 m
Reference depth initial reservoir pressure	≈396 bar
Reservoir temperature	≈147 °C

The saturation relations in the reservoir are of particular interest as the reservoir fluid is a gas condensate, containing around 50% of CO₂ in the composition. Numerous phase changes occur in the system (surface installations, wells, and reservoir), due to changing p-T conditions during CO₂ injection. This is why compositional simulation was inevitable, as it accounts not only for the natural gas composition, but also the phase changes (near production wells, the gas, mostly CO₂, condensates in later production periods when the pressure falls below the critical value), which gives a fairly realistic representation of the saturations and distribution of the reservoir and injected fluid individually, including their mixture.

2.5. History Matching and Base Case Prediction Scenarios

The efficiency of each injection strategy could only be assessed if there are baseline scenarios that they can be compared to [46]. Each of the three simulated base cases consists of three characteristic phases:

• Primary production based on historical data, used for assessment of the model reliability (history matching, HM).
• Extrapolated production with identical production constraints as in HM (business as usual, BAU).
• ‘Pure’ storage after the reservoir depletion (end of production).

The model was validated by simulating the first 16.5 years of production and comparing it to the actual historical data. In all three base scenarios (that differ by the selection of the production wells), an acceptable history matching of cumulatively produced gas was obtained (Figure 5). To achieve this, the production rates were adjusted to obtain the material balance that was observed in this reservoir in reality.

Figure 5. Comparison of cumulative production for the base cases and actual data.

In each case, two production wells were selected for simulating the overall reservoir production, while the third one was only activated for the injection after the reservoir depletion. The reason for not including all three wells in the storage period is to obtain the identical well pattern in the base case and the corresponding injection cases (making the scenarios comparable).

2.6. The Injection Scenarios Definition

After history matching individual base scenarios, they were used for building injection (prediction) scenarios that differ from each other by the start of the injection (Table 3).

Table 3. Timings of the CO₂ injection scenarios.

	Nomenclature of the Cases
Start of the Injection, Years After the Start of Production	Injection Well
	B-9	B-8	B-6
0	CCS1_2000	CCS2_2000	CCS3_2000
5	CCS1_2005	CCS2_2005	CCS3_2005
8	CCS1_2008	CCS2_2008	CCS3_2008
16	CCS1_2016	CCS2_2016	CCS3_2016
20	CCS1_2020	CCS2_2020	CCS3_2020
25	CCS1_2025	CCS2_2025	CCS3_2025
30	CCS1_2030	CCS2_2030	CCS3_2030
35	CCS1_2035	CCS2_2035	CCS3_2035
40	HM1_BAU_STORE	HM2_BAU_STORE	HM3_BAU_STORE

Following the evaluation of strategies, including the injection well bottom-hole (dynamic) pressure limitations, the decision was made to adopt the method of volumetric replacement at the reservoir level during the concurrent natural gas production and CO₂ injection. The main reason for this decision is to alleviate an excessive reservoir pressure increase. Production rates were determined to match the base cases’ production. In all cases, hydrocarbon production lasts for 40 years, after which the ‘pure’ storage phase was simulated for 60 years. The injection rate in the storage period was 150,000 m³/day (at the surface, unlike the reservoir rate during the period of injection and simultaneous production), which is considered to be realistic for one well in a depleted reservoir in this specific field, considering the historic data.

3. Results and Discussion

Considering different aspects from which the simultaneous natural gas production and CO₂ injection can be viewed, several indicators were observed to assess the success of each strategy: cumulative gas production (hydrocarbons without CO₂), cumulatively produced CO₂, retention, storability (storage efficiency), and reservoir pressure.

3.1. Production Indicators

Cumulative hydrocarbon production was compared for all prediction scenarios (Figure 6, Figure 7 and Figure 8), to determine the CO₂ breakthrough and the original reservoir fluid contamination and to evaluate the success of the scenarios, considering the increase in hydrocarbon production. This is especially interesting, considering that the original reservoir fluid has a high CO₂ content.

Figure 6. Cumulative hydrocarbon production for the case with B-9 as the injection well.

Figure 7. Cumulative hydrocarbon production for the case with B-8 as the injection well.

Figure 8. Cumulative hydrocarbon production for the case with B-6 as the injection well.

For cases where injection wells are B-8 and B-9 (strategies CCS1 and CCS2), the base-line scenarios (BAU) have the highest cumulative hydrocarbon gas (without CO₂) production, with the exception of the latest timing (35 years after the production start). This can be attributed to minimal fluid contamination. However, for the scenarios where the injection starts simultaneously with production (CCS1_2000 and CCS2_2000, and also CCS3_2000), hydrocarbon gas production in the first fifteen years is greater than the corresponding base cases. Cumulative production at the end of the observed period is still lower compared to the base cases. This can be attributed to the earlier breakthrough of CO₂. Nevertheless, cumulative productions that are similar to BAU scenarios are noticed in scenarios where the injection is started only after the reservoir depletion (CCS1_2030 and CCS1_2035, and CCS2_2030 and CCS2_2035). This is in line with the findings from the literature [25,46], where the simulations implied that it is desirable to deplete the reservoir as much as possible prior to the implementation of EGR (when the goal is to maximize natural gas production, i.e., recovery).

The same phenomenon is observed in the third well arrangement (B-6 as the injection well), but in these strategies, more cases achieve higher cumulative hydrocarbon production than the base case. This can be attributed to the injection and production wells’ selection. When B-6 (as the deepest of the three wells) is chosen as the injector, hydrocarbon production is lower than in the other two well positionings. This indicates that B-6 has the best productivity (cumulative hydrocarbon production is higher if B-6 is chosen as a producer), and the best injectivity too—the most favorable retention results are achieved with well B-6 as the injector, which implies that this well is positioned in the most favorable part of the anticline. Further EGR optimization could be achieved by varying the production and injection rates.

A later injection start increases the ultimate recovery, but for storage assessment, the effect on the CO₂ retention efficiency must be taken into account as well. Although the ultimate recovery is a criterion that affects the project’s profitability, other parameters have to be considered to assess the suitability of the reservoir for storage with simultaneous production.

Cumulative CO₂ production has the same trend, regardless of the wells’ arrangement (Figure 9). The base cases produce the smallest amounts of CO₂ because CO₂ is not injected, and the only CO₂ production is from the original reservoir fluid. The greatest CO₂ production is obtained in cases where the injection starts early, meaning that the fluid composition degradation (CO₂ content increase) cannot be avoided. However, it can be optimized by varying the production and injection rates.

Figure 9. Cumulative CO₂ production.

3.2. Retention

As opposed to the previously evaluated criteria that are relevant only in the production phase (the first 40 years), retention (Figure 10) should also be observed in the ‘pure’ storage period (the last 60 years). A positive value of retention indicates that CO₂ is permanently stored (more CO₂ is injected than is produced) and also confirms gate-to-gate carbon negativity of the process.

Figure 10. Retention results.

Retentions follow the same trend in all strategies, but their values vary because of the different quantities of the produced and injected CO₂. All scenarios that include both injection and production have positive retentions from the start of injection.

When observing earlier injection starts, the CCS3 scenarios achieve notably lower retentions than cases with alternate well arrangements. For later injection starts, this effect changes, and the latest injection start yields the highest retention in CCS3 strategies. Furthermore, CCS1 scenarios only yield the highest retention in the first two injection timings (compared to the other two well arrangements), and it can be concluded that this well placement is the most sensitive to the injection start, since the retention for this set of scenarios falls below retentions of the other two well arrangements in later injection starts. Considering that reservoir properties (namely porosity, which, in combination with permeability and the productive interval thickness, dictates the well’s deliverability) around the injection well (B-9) are poorer compared to the producers (B-6 and B-8), the described effect can be explained by lower injectivity than productivity, which becomes more prominent with time.

When comparing individual cases for each well’s arrangement, the highest retention at the end of the observed period is (in general) achieved in those where the injection starts earlier. This is explained by the highest cumulative amounts of CO₂ injected. The only exceptions were observed in strategies CCS2 and CCS3, when comparing scenarios with the injection starting in the eighth year (after the start of production) to scenarios with the injection starting in the fifth year (retention in CCS2_2008 is higher than in CCS2_2005, and retention in CCS3_2008 is higher than in CCS3_2005).

The ultimate retentions in the baseline scenarios are low, leading to the conclusion that simultaneous CO₂ injection and gas production increase the CO₂ storage capacity of gas reservoirs.

Negative retention values mean that less CO₂ is injected than is produced; therefore, the case is not carbon negative and, considering the high initial CO₂ content in the reservoir fluid, negative retention is expected. This implies that the only way to achieve carbon negativity in this specific field is to inject CO₂.

3.3. Storability

Another important indicator that has to be observed in both production and ‘pure’ storage phases is storability.

In the production period of CCS1 scenarios (Figure 11), better results for storability were generally observed for an earlier injection start, which is in accordance with retention results for these cases. It was noticed that storability decreases with time in the production phase, and this can be attributed to an increase in CO₂ breakthrough.

Figure 11. Storability for cases with B-9 as the injection well.

Somewhat different storability results are observed in CCS2 (Figure 12) and CCS3 (Figure 13), as later injection starts yielded initially higher storability values.

Figure 12. Storability for scenarios with B-8 as the injection well.

Figure 13. Storability for scenarios with B-6 as the injection well.

The ‘pure’ storage phase scenarios with later injection starts yielded the best storability results. This can be attributed to a fast decrease in CO₂ production and a somewhat slower increase in the retention. The highest storability for all wells’ arrangements (CCS1, CCS2, CCS3) is achieved in scenarios with the latest injection start, and it is significantly higher than in cases with earlier injection starts.

The differences in the results can again be attributed to the well positioning—the scenarios with the deepest well (CCS3) as the injector yield higher storability values, meaning that the amounts of CO₂ injected exceed those of the CO₂ produced (the result of leaving out the “best” well from the production period and activating it in the injection period), and their difference is favorable in relation to the cumulatively produced CO₂. Aside from the well depth (which dictates the pressure, i.e., reservoir energy), other important factors influencing this well’s best performance are the reservoir properties, such as porosity, permeability, and productive interval thickness, which are all more favorable in the vicinity of B-6 compared to the other two wells.

During the storage phase of the base cases, there is no production; hence, all CO₂ injected in these scenarios will be permanently kept (Figure 14). It is evident that the volume of injected CO₂ is consistent throughout all base cases, mostly because of the selected injection restriction of 150,000 Sm³/day. Low storability values of the base case scenarios do not indicate poor storage performance, but are a result of the very long period without any injection whatsoever. It is practically impossible to store a sufficient CO₂ quantity at the given conditions (namely the injection rate) during the storage phase (60 years) to offset all CO₂ produced in the production phase (40 years).

Figure 14. CO₂ injection, cumulative for the scenarios with ‘pure’ storage after the end of production.

3.4. Pressure Behavior

The retention amounts for the base cases (shown in Section 3.2) do not represent the ultimate storage capacities, as the reservoir pressure is not restored in 60 years of injection (Figure 15), meaning there is still available pore space for further CO₂ injection (storage).

Figure 15. Average reservoir pressure for the scenarios with ‘pure’ storage after the end of production.

However, it was out of the scope of this work to investigate the influence of different injection rates in the ‘pure’ storage period. This would surely change the dynamics of the reservoir pressure increase in time.

Some of the cases yield an increase in reservoir pressure above the initial almost immediately as the ‘pure’ storage phase begins, but in general, it can be concluded that the reservoir can be used as a CO₂ storage facility for a long time (Figure 16). Throughout both the production and injection phases, pressure typically remains above its critical value (73.8 bar), meaning that the reservoir process is almost always supercritical.

Figure 16. Reservoir pressure for cases with simultaneous production and injection.

Aside from the reservoir (or productive interval) thickness, reservoir pressure is also an important factor influencing gravity segregation, as it dictates the CO₂ density. Under given reservoir p-T conditions, it can be expected that CO₂ would have a significantly higher density than the original reservoir fluid, which enhances piston-like displacement.

The reservoir’s capacity, i.e., CO₂ storage potential, can be assessed by analyzing the pressure–retention relationship for each scenario. Because of the high CO₂ content in the original fluid composition, BAU_STORE scenarios yield decreasing retention until the CO₂ injection starts. The retention in these cases stays negative for most of the ‘pure’ storage phase, i.e., the period in which the reservoir pressure increases (Figure 17).

Figure 17. Retention vs. pressure for all cases with simultaneous production and injection.

In the scenarios with an early injection start, the reservoir is not sufficiently depleted and the pore volume available for further injection is rather low. Higher initial retention in earlier injection timings (as illustrated in Figure 10 and Figure 17) can mostly be attributed to the highest cumulative CO₂ injection leveraging the large amounts of CO₂ produced in these cases. However, the ultimate storage capacity is rather limited, considering the reservoir repressurization—the initial reservoir pressure is reached at relatively low retentions (less than 6 × 10⁹ m³). The main justification for an early injection start can be found in favorable discounted cash-flow, resulting from produced gas sales. Additionally, an early injection start eliminates the need for capital investments at a later ‘pure’ storage phase. Extrapolation of the lines, under the assumption that they would exhibit the same trend in the future, leads to the conclusion that higher ultimate retention could be achieved in ‘later’ scenarios, since the reservoir pressure is far from the initial. Extrapolation of the BAU cases shows that they would not always yield the highest ultimate retention, considering the initial reservoir pressure. It should be noted, however, that in practice, the storage design would include far more consideration of various injection rates.

Finally, the scenarios can be compared based on the CO₂ content in the produced gas (Figure 18).

Figure 18. CO₂ mole fractions.

It can be observed that in earlier injection timings, the CO₂ content eventually exceeds 85% (and even 90% in some cases), which might imply that these scenarios are less feasible compared to those with later injection starts. It can also be concluded that well placement has a notable influence on the CO₂ content, as the scenarios with B-8 as the injection well (CCS2) have the highest CO₂ fraction, no matter the injection timing.

However, the absolute values of CO₂ content in the produced gas at the end of the observed period cannot be the only criterion for scenario assessment, since the scenarios with a late injection start have less time to exhibit the CO₂ content trend to its fullest extent. It can be seen that the CO₂ fraction in the CCS2_2030 scenario exceeded 80% less than 10 years after the injection start. It is therefore useful to detect the times of significant increase in the CO₂ fraction for each scenario (Table 4).

Table 4. Time of significant increase in CO₂ content.

Scenario	Injection Year	Year (Since the Production Start) of the Observed Change in CO2 Content (Relative to the Original CO2 Fraction)
		10% Change	20% Change	30% Change
CCS1	2000	5.55	7.41	9.28
CCS1	2005	12.14	14.82	17.41
CCS1	2008	13.28	15.27	17.28
CCS1	2016	20.14	21.69	23.41
CCS1	2020	23.55	24.91	26.28
CCS1	2025	28.82	30.41	31.91
CCS1	2030	33.82	35.27	36.82
CCS1	2035	38.69	40.14	-
CCS2	2000	4.68	6.34	8.41
CCS2	2005	10.91	13.00	15.55
CCS2	2008	12	13.41	15.14
CCS2	2016	18.91	20.14	21.41
CCS2	2020	22.57	23.62	24.68
CCS2	2025	27.00	27.82	28.82
CCS2	2030	32.41	33.28	34.28
CCS2	2035	37.82	38.91	40.14
CCS3	2000	8.91	11.55	15.41
CCS3	2005	16.14	20.82	26.28
CCS3	2008	15.82	19.41	23.27
CCS3	2016	20.27	22.28	24.27
CCS3	2020	24.68	26.91	29.14
CCS3	2025	30.00	32.41	34.82
CCS3	2030	35.00	37.41	39.91
CCS3	2035	40

CCS3 scenarios, where the injection well is the one with best flow properties (B-6), exhibited delayed CO₂ breakthroughs. This can be ascribed to relatively low productivity of the two producing wells, and is in line with storability results, which showed that this well arrangement yields the best ratio of retained (permanently stored) and produced CO₂.

Figure 19 is a graphical representation of these data, and it can be concluded that the earliest injection timing has a moderate increase in CO₂ fraction, while the injection start in the fifth year (from the production start) exhibits the biggest delay in the CO₂ fraction increase, no matter the well placement.

Figure 19. Occurrence of significant increases in CO₂ fraction.

Although some similarities in trends for all well placements were detected, the CO₂ breakthrough cannot be unambiguously determined, which can be mostly attributed to the process miscibility.

4. Conclusions

Suitability of a reservoir for permanent CO₂ storage can be estimated through a set of indicators, and by strategically positioning the injection well and determining the optimal injection start time, it is possible to optimize CO₂ storage strategies according to the retention criterion. The results drawn from the analysis of simulation strategies are as follows:

• An earlier start of CO₂ injection results in greater retention (i.e., CO₂ storage) at the end of the observed period, no matter the well pattern. However, those ‘early’ scenarios reach the initial reservoir pressure soon after the ‘pure’ storage period starts, meaning that, at given production and injection rates, the ultimate retention is lower compared to later injection starts. This is partly a consequence of increased gas (including CO₂) production in the period of simultaneous injection and production.
• All CCS scenarios become (gate-to-gate) carbon negative almost immediately at the start of injection, and only one of the base cases resulted in negative retention value (more CO₂ was produced than injected), but this is understandable, considering the fraction of CO₂ present in the initial reservoir fluid composition.
• The importance of well placement became obvious when the well with the most favorable flow properties was selected as the injector. These scenarios exhibited the best storability and the latest CO₂ breakthrough, meaning that properties such as reservoir porosity, permeability, and productive interval thickness, along with the reservoir pressure, significantly influence the storage efficiency, probably due to gravity segregation and miscibility effects.
• Simultaneous gas production and CO₂ storage could be feasible in this specific reservoir, because of the readily available CO₂ (from this reservoir but from nearby reservoirs as well), which is the primary carrier of costs in EGR projects.

The main factors influencing the optimal strategy are technological limitations and the ultimate goal (e.g., increased recovery or efficient storage). Later injection timings result in slightly higher hydrocarbon recovery, which can be partly attributed to lesser fluid contamination. The fast increase in reservoir pressure in the storage phase that occurred in cases with early injection supports the conclusion that a high-CO₂-content gas reservoir should be at least partly depleted before the injection start. Further optimization would be achieved by changing CO₂ injection rates. In the region where the analyzed gas reservoir is located, it is common to find gas compositions containing approximately 50% or more CO₂. The facilities for gas processing (purification) are constructed before the discovery of this particular gas reservoir, resulting in a favorable economic aspect for gas processing. In gas reservoirs that have a low initial CO₂ content, the separation of produced CO₂ following the CO₂ breakthrough will show up as a significant economic factor. In such cases, increased CO₂ retention can be expected, significantly modifying the economics of the entire concept. This indicates the need for further research focusing on the initial CO₂ content and the economics of the complete gas production and CO₂ (re)injection cycle.

No investments in new wells were considered, as all options analyzed in this paper count on the existing infrastructure (wells and technology).

The results presented here include the CO₂ flow simulation in a complex (heterogeneous) model of an existing producing gas reservoir. Although the composition and properties of the reservoir fluid are very unfavorable (around 50% mol of CO₂, which is not unusual for the observed area), gate-to-gate (injected CO₂ from capture and production wells vs. CO₂ at production wells) carbon negativity can be achieved by the selection of appropriate timing and injection well. It is expected that the results would be even more favorable if the initial composition contained less CO₂. Additionally, scenarios with simultaneous CO₂ injection/storage and production are considered justified, as this concept creates market value. Furthermore, CCS technology is readily available, thus eliminating the standard estimate issues, where the injection starts up to five years after the hydrocarbon reservoirs depletion. Finally, regarding CO₂ storage in deep saline aquifers, it is expected that the first projects could start within ten years of the initial storage capacity assessments.