FIELD OF THE INVENTION

The present invention relates generally to carbon sequestration, and specifically to subterranean carbon sequestration technologies.

BACKGROUND OF THE INVENTION

Geological carbon storage is part of Carbon Capture and Storage (CCS) technologies, where industrial CO2 is captured at its source, transported and permanently stored in the deep subsurface. Any CO2 migration outside of a subsurface geological storage reservoir is a main concern, and site operators need to ensure no significant risk of leakage of CO2 will occur over thousands of years. Managing the risk of leakage involves subsurface monitoring and being able to respond in the event of CO2 migration through compromised cap rocks, faults/fractures or impaired well bores.

CO2 storage sites typically consist of a reservoir with high porosity and permeability and a cap rock above the reservoir with low porosity and permeability. Such sites have little pressure build-up during the injection and the cap rock ensures containment. However, such reservoirs are known to utilise only a very small fraction of the available pore space, as buoyancy-driven CO2 transport dominates and most CO2 accumulates as free CO2 over a large area underneath the cap rock. CO2 plume stabilisation can take a long time and monitoring of a large area is generally required.

Enhanced pore space utilisation approaches have been proposed, but remain untested under in-situ conditions. In addition, several technologies and materials have been considered with the aim to form barriers that would block the leaking pathways of pre-existing CO2 reservoirs, such as the use of cements and geopolymers. However, their applicability is only effective very close to the wellbore due to their high viscosity, inherently posing risk of blockage. Bio-mineralization has also been proposed as a leakage mitigation technology in pre-existing CO2 reservoirs, for example by using microbes to promote calcite precipitation to seal leaks. It has also been proposed to contain CO2 leakage temporarily using hydraulic barrier formation, which involves water injection into aquifers overlying the caprock and thereby redirecting the CO2 flow direction. Most of these technologies consider the pressure and temperature under CO2 storage conditions but may be unsuitable for use under these mildly acidic environments.

There remains therefore an opportunity to address or ameliorate limitations associated with current carbon dioxide sequestration technologies.

SUMMARY OF THE INVENTION

The present invention provides a method of forming a subterranean silica-gel flow diverter in a carbon sequestration reservoir, the method comprising a step of:

(a) co-injecting an acidic solution and an alkali metal silicate solution at a first depth below a first storage volume of the reservoir, such that the solutions mix while distributing laterally; or

(b) mixing an acidic solution and an alkali metal silicate solution and injecting resulting mixed solution at a first depth below a first storage volume of the reservoir, such that the mixed solution distributes laterally; wherein silica-gel forms as the co-injected solutions mix or from the mixed solution to provide said flow diverter, which is configured such that at least some of carbon dioxide injected at a depth below said flow diverter is directed laterally within the reservoir before entering the first storage volume.

The proposed method presents itself in direct conceptual contraposition to existing mitigation technologies. Traditional mitigation procedures deal with existing CO2 reservoirs and are aimed at blocking leakage of CO2 that has already been injected underground. In contrast, the method of the invention presents as a “pre-conditioning” procedure performed on a suitable reservoir volume before CO2 is injected. The co-injection of an acidic solution and an alkali metal silicate solution or injection of a mixture of an acidic solution and an alkali metal silicate solution (mixed solution) results in the generation of a subterranean silica-gel deposit (flow diverter), which is typically disc-shaped. Said flow diverter can improve the subterranean distribution of subsequently injected CO2 over a large underground volume by either substantially prohibiting direct vertical spread of the CO2 plume and spreading it laterally before it enters the storage volume (in the case of formation of a flow diverter that is substantially CO2 impermeable) or by reducing and/or slowing the rate of direct vertical spread of the CO2 plume and spreading it to some extent laterally before it enters the storage volume (in the case of formation of a flow diverter that is semi- permeable and allows CO2 permeability to at least some extent). Without wanting to be limited by theory, it is postulated that the subterranean silica-gel deposit functions as flow diverter by altering the porosity characteristics of the subterranean injection site, such that direct vertical CO2 flow can be significantly minimised or at least slowed or reduced, in favour of lateral migration.

Throughout this document reference to "injection" of acidic solution and an alkali metal silicate solution is intended to encompass both co-injection of acidic solution and an alkali metal silicate solution and injection of a mixed solution of acidic solution and an alkali metal silicate solution.

Since the injection is effected at an injection point below a storage volume of the reservoir, the silica-gel flow diverter forms below said storage volume. This ensures that CO2 subsequently injected at an injection point below the flow diverter can spread laterally along the flow diverter until it reaches its boundaries. At that point, buoyancy-driven vertical transport of CO2 can commence over a significantly enlarged subterranean portion of the reservoir relative to the injection point, ensuring homogeneous and effective CO2 distribution into the reservoir and consequent enhanced trapping in the reservoir due to prevailing capillary forces. Overall, the proposed silica-gel flow diverter can advantageously reduce vertical mobility of injected CO2 through optimised subterranean lateral distribution and may result in improved CO2 storage capacity and/or efficiency. A schematic representation of one embodiment of the proposed flow diverter (i.e. the mixed solution injection aspect) is shown in Figure 1. Co-injection of acidic solution and an alkali metal silicate solution is effected in much the same way as the injection depicted in Fig. 1, except that in that case rather than the acidic solution and an alkali metal silicate solution being premixed at the surface and then injected as a mixture to the injection site beneath the storage volume but above the CO2 injection location, they are injected as separate solutions at the same site, where they move laterally and mix.

Silica-gel can form by condensation of hydrolysed silicates under pH neutral and acidic conditions. As the acidic and alkali metal silicate solutions are co-injected underground, the solutions mix and preferentially spread laterally resulting in formation of a silica-gel flow diverter that is typically disc-shaped. That is, the flow diverter has a large lateral to vertical ratio, such as for example from about 10:1 to about 100:1, such as about 15:1 to about 80:1, about 20:1 to about 50:1, or about 25:1, about 30:1 or about 40:1.

Advantageously, the co-injection of the two solutions can ensure fine control over the kinetics of the silica-gel formation. In turn, this can assist with formation of a homogeneous flow diverter in the subterranean volume of interest. The rate of silica gel formation is primarily dependant on the pH of the final solution, which can be controlled through the composition and mixing of the two precursor solutions. Therefore, co-injecting the precursor solutions ensures controlled precipitation of silica-gel relative to sequential injection of the two solutions.

Conversely, in the case where the acidic solution and an alkali metal silicate solution are pre-mixed and then injected as a mixed solution there may be an advantage in terms of the extent of mixing of the solutions. Whereas in the case of co-injection of the two solutions the gelation reaction will take place at the interface between the two solutions, in the case of pre-mixing and injection of the mixed solution the two solutions will likely be more intimately mixed, resulting in more extensive gelation taking place. The rate of gelation can be controlled to ensure that extensive gelation dos not occur prior to completion of injection by tuning the characteristics of the solutions (as discussed in further detail below), to ensure that the injection at the injection site is not inhibited as a result of undue mixture viscosity. The silica-gel flow diverter is intended to improve the efficiency of carbon sequestration within the reservoir volume of interest. Accordingly, the method of the invention additionally comprises the injection of carbon dioxide at a depth below the flow diverter. In some embodiments, the carbon dioxide is injected at an injection point which is in close proximity to the flow diverter, for example 2 - 10 meters below the flow diverter. This advantageously ensures improved lateral migration of the carbon dioxide below the diverter, resulting in more homogeneous distribution of the gas before it enters the reservoir.

It can be advantageous to inject CO2 after an incubation period following the injection of the acidic solution and the alkali metal silicate solution or mixed solution. Said incubation period can be beneficial to increase the overall amount of silica-gel that forms, thereby ensuring that silica-gel can effectively form homogeneously across the entire subterranean injection volume. Accordingly, in some embodiments the method comprises injecting CO2 at a depth below the flow diverter after an incubation period of at least 24 hours. In some embodiments, CO2 is injected after an incubation period of at least 5 days.

It is postulated that the method of the invention will contribute to a more efficient, costsaving and safer geological CO2 storage. This can be achieved by drastically minimising mobility of the CO2 within the underground reservoir, resulting in a larger proportion of CO2 being no longer stored in form of a mobile plume, as in conventional storage procedures. The method of the invention can advantageously improves capillary trapping of CO2, with the result that free, mobile CO2 is much reduced relative to conventional sequestration procedures.

The present invention also relates to a subterranean silica-gel flow diverter per se, said flow diverter extending laterally in a carbon sequestration reservoir at a first depth below a first storage volume of the reservoir, and wherein said flow diverter is configured such that carbon dioxide injected at a depth below said flow diverter is directed laterally within the reservoir before entering the first storage volume. Said flow diverter may be one that has characteristics described herein, for example one that is obtained in accordance to the method described herein. BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will be now described with reference to the following nonlimiting drawings, in which:

Figure 1 shows a schematic representation of the effect of the flow diverter on the distribution of subsequently injected CO2 with an impermeable barrier (a) and a semipermeable barrier (b).

Figure 2 shows the speciation of silica as a function of pH at a) 25°C and b) 60°C. The stability field of quartz, chalcedony, trydimite and cristobalite have been suppressed to reveal the stability field of amorphous silica (Si02(am)), which preferentially forms as an amorphous solid or gel depending upon silica concentration and pH, and

Figure 3 shows spatial distribution of relevant parameters relative to the flow diverter after 5 days of incubation, namely: a) pH, b) amount of silica precipitated, c) porosity, d) permeability, e) H4(H2SiO4)4 ^-4 , f) H6(H2SiO4)4 ^-2 , and parameters after 4 days of injecting CO2: g) Sg, h) pH, i) porosity, j) permeability. The X-axis is 50 m long.

Figure 4 shows changes in Si02(am) abundance along the flow-through column where x = 0 cm is the inlet. Injection of CCh-enriched water into a Na-Si saturated column with continuous injection for 24h (A) and with injection / no-flow cycles [20min inj, 8h n-f; 4min inj, 16h n-f; 4min inj, 16h n-f; 7.53h inj] for a total of 48h (B). Si02(am) precipitation was modelled with a molar volume of 29 cm ³ /mol.

Figure 5 shows distribution of pH (A), Si02(am) abundance (B), the Si02(am) saturation index (C) and porosity (D) in the flow-through column at different times during the intermittent injection of CCh-enriched water into a Na-Si saturated column. X is the distance from the inlet. Injection sequence: 20min inj, 8h n-f; 4min inj, 16h n-f; 4min inj, 16h n-f; 7.53h inj. Si02(am) precipitation with a molar volume of 1000 cm ³ /mol. Figure 6 shows a conceptual model of the silica-gel flow diverter.

Figure 7 shows results of near wellbore simulation (CASE A) where 1) the reservoir was saturated with the alkaline Na-Si solution, 2) scCCh was injected and 3) a slug of water was injected. X-axis with a length of 50 metres.

Figure 8 shows results of near wellbore simulation where 1) the reservoir was saturated with the alkaline Na-Si solution, 2) water was injected for 3 (Case B) and 6 (Case C) hours and 3) scCCh was injected for 12 hours following the water injection. X-axis with a length of 5 metres.

Figure 9 shows results of Cased D: a mitigation scenario where groundwater was injected for 3 hours followed by 1 week of scCCh injection. X-axis with a length of 50 metres. A: pH distribution, B: Si02(am) distribution, C: porosity distribution.

Figure 10 shows results of a mitigation scenario where groundwater was injected for 3 hours followed by 1 week of scCCh injection. 3 months after shut-down injection. X-axis with a length of 50 metres. A: pH distribution, B: Si02(am) distribution, C: porosity distribution.

Figure 11 shows results of a mitigation scenario where groundwater was injected for 3 hours followed by 1 week of scCCh injection. 6 months after shut-down injection. X-axis with a length of 50 metres. A: pH distribution, B: Si02(am) distribution, C: porosity distribution.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to a method of forming a subterranean silica-gel flow diverter.

Reference herein to a subterranean “flow diverter” means reference to a subterranean volume that offers reduced permeability to CO2 relative to its surroundings. The reduced permeability to CO2 offered by the flow diverter forces CO2 injected below said flow diverter to migrate laterally along the lower boundary of the diverter, as opposed to flowing vertically through the diverter. As the CO2 reaches the lateral boundaries of the diverter, buoyancy- driven vertical transport of CO2 into the storage volume above can commence. Figure 1 shows a schematic representation of an embodiment of the working mechanism of the proposed flow diverter.

Advantageously, the provision of a flow diverter in accordance to the method of the invention improves the carbon sequestration storage capacity of the reservoir volume above the diverter. Without wanting to be limited by theory, it is postulated that the enlarged base for vertical buoyancy-driven CO2 plume offered at the boundaries of the flow diverter is effective to slow down the overall vertical CO2 flow, giving rise to a significantly large lateral CO2 transport and favouring trapping by capillary forces within the reservoir.

In essence, it is believed that the proposed flow diverter acts to slow the vertical buoyancy- driven CO2 plume and spread it over a larger lateral volume relative to the injection point, maximising its distribution and sequestration within the reservoir by capillary forces. The proposed flow diverter can therefore afford a more efficient use of the pore space within the reservoir and a larger proportion of CO2 will be stored by capillary trapping as opposed to merely storing free, mobile CO2 underneath a cap rock. The method of the invention can therefore afford significantly larger CO2 storage capacity for underground reservoirs even in the absence of a defined cap rock.

By being a “silica-gel” flow diverter, the flow diverter of the invention is essentially made of silica-gel. As it is known in the art, silica-gel is an amorphous form of silicon dioxide (S i O2) consisting of an irregular tridimensional framework of alternating silicon and oxygen atoms with nanometer-scale voids and pores. As a skilled person would know, major solid phases of silica or silicon dioxide (SiCh) are generally a) amorphous, un-crystallized and optically opaque silica-gel or opal, b) microcrystalline chalcedony, and c) well crystallized, optically anisotropic quartz.

The stability of the aqueous form of silica is dependent on pH and temperature. This is illustrated by the phase diagrams shown in Figure 2, depicting the speciation of silica as a function of pH at 25°C (Figure 2(a)) and 60°C (Figure 2(b)). The simplest soluble form of SiCh, is the orthosilicic acid monomer (H4SiO4, represented by the SiO2(aq) species in Figure 2), which is universally found in seawater, fresh waters and the interstitial waters of soils and rocks up to a pH of around 9.8 (see Figure 2(a)). Monomers of silicic acid can combine to form larger oligomers, initially a dimer (two monomers bound together), which can polymerise forming larger molecules. In the case of silica, said polymerisation is a polycondensation reaction whereby water is released. These species eventually aggregate as gels depending on the presence of salts or other charged species present. The polymerization of silica at sufficiently high concentrations eventually leads to the precipitation of an amorphous solid (SiO2(am)) gel under certain conditions of pH and/or silica concentration (see e.g. Figure 2). These precipitates are stable under pH neutral and acidic conditions and may therefore act as a permeable or semi-permeable barrier to flow through a porous media for a long time.

Accordingly, in the context of the invention the silica-gel flow diverter is made of amorphous silica resulting from the poly-condensation reactions of silanol groups (-Si-OH) of silicate species in solution. Said poly-condensation reaction will be effected at conditions (e.g. pH, temperature, and silica concentration) at which amorphous silica forms as a stable phase. The present invention has been developed following the observation that SiO2(am)-gel can be effectively implemented as a tool for generating CCh-impervious subterranean volumes or subterranean volumes that at least slow or to some extent restrict CO2 transition, that can act as flow diverters for the gas.

In the context of the invention, it will be understood that operative conditions of the method of the invention (e.g. nature and amount of acidic and alkali metal silicate solutions, injection depth, injection temperature, injection rate, injection hydrostatic pressure, mixing conditions etc.) will be devised to ensure effective subterranean formation of SiO2(am)-gel suitable for the intended purpose. In that regard, a skilled person will be readily capable to find suitable guidance from information such as that provided by phase diagrams of the kind shown in Figure 2. Such information will guide a skilled person to adopt operative conditions that are ultimately favourable to ensure that poly-condensation reactions resulting in silica-gel will be promoted under conditions (e.g. pH, temperature, time, silica concentration, etc.) at which amorphous silica-gel forms as a stable phase. For instance, the injection step (including coinjection) in the method of the invention will be effected to ensure that, for a given temperature, the resulting pH of the underground volume of interest are suitable for the formation of amorphous silica-gel as the stable silica phase.

Accordingly, the injection can be effected to achieve any pH suitable to achieve formation of silica-gel as the stable silica phase. Typically, the injection step in the method of the invention achieves a pH of about 11.5 or less, about 11 or less, about 10.5 or less or about 10 or less in the injection volume. In some embodiments, the injection step in the method of the invention achieves a pH of about 9.5 or less, about 9 or less, about 8 or less, or about 7 or less in the injection volume.

It is possible to tune the precipitation rate of the silica-gel by tuning the pH of the subterranean injection solution. As a general principle, the lower the pH the faster the silica- gel tends to precipitate. A skilled person would be readily capable to identify adequate pH values of the underground volume of interest to obtain a suitable precipitation rate of the silica-gel. For example, tables correlating the pH to the precipitation rate of the silica-gel are readily available in the art.

As noted above the silica gel forms from solution over time and the rate of gel formation is generally dependent on pH, as well as silica concentration, temperature and cation concentrations in the mixed solution containing the alkaline silicate solution and an acid. The rate of gel formation can be estimated based on empirical data. The gelation time becomes shorter with a lower pH, a higher silica concentration and a higher temperature. For example, the gelation time, t _g , for a gel not passing through a filter with a pore size of 3 micrometres equivalent to a turbidity of 170 NTU can be determined according to Stavland et al. ^16-18 where [Si] is the silicate concentration, [HCl]x is the effective acid concentration accounting for the silica concentration, [Ca ²⁺ ] is the calcium concentration in the makeup water, and Ea/RT is the Arrhenius type thermal energy. The exponents a, P and y are determined from experiments.

For example, the gelation time (tg) is reached for a given solution with a pH of 10.9 and a temperature of 40 °C after one day, a solution with a pH of 11.15 and a temperature of 40 °C after 10 days. A gelation time (tg) is reached for a given solution with a pH of 10.9 and a temperature of 60 °C after 0.8 days, a solution with a pH of 11.3 and a temperature of 60 °C after 10 days ¹⁶ .

The pH of the underground volume of interest can also affect the porosity of the resulting silica-gel. Typically, the lower the pH the lower the porosity. Without wanting to be limited by theory, it is postulated that lower values of pH result in higher concentration of ionised silanol groups. Said groups are mainly present as Si-OH groups that can readily undergo polycondensation into Si-O-Si groups. This principle may offer guidance to the skilled person to determine a target pH for effective formation of silica-gel having porosity characteristics that suitable to divert CO2.

The method of the invention involves forming a subterranean silica-gel flow diverter in a carbon sequestration reservoir. In the context of the invention, a “carbon sequestration reservoir” will be understood to be a subterranean volume having geologic characteristics (such as depth, chemical composition, porosity, etc.) making it suitable to entrap CO2. Without wanting to be confined by theory, CO2 entrapment may be achieved by any trapping mechanism, for example residual trapping due to capillary forces, dissolution trapping, and mineral trapping. Examples of suitable carbon sequestration reservoirs include depleted oil and gas fields and underground saline formations. These are known to comprise layers of porous rock (e.g. sandstone) typically deeper than 800 meters below surface, which may be located underneath a layer of impermeable rock (cap-rock). Typically, suitable carbon sequestration reservoirs in the context of the invention will be those that present conditions suitable for the sequestration of CO2 as a supercritical fluid. In some embodiments, the carbon sequestration reservoir is located at a depth of at least 800 metres.

The method of the invention comprises a step of either co-injecting an acidic solution and an alkali metal solution or injecting a mixture of these two solutions.

By “co-injecting” the solutions is meant that at least a portion of the acidic solution and at least a portion of the alkali metal solution are injected at the same time at said first depth below said first storage volume of the reservoir. As a result, the co-injected solutions mix at said injection depth and can travel laterally as a mixture. Said mixture will have a pH conducive to precipitation of silica-gel as a stable phase. Alternatively, the solutions can be pre-mixed, for example at the surface, and the mixture is then injected. Mixing of the solutions can be achieved using conventional means, for example using a turbine or impeller.

By the solutions being co-injected “at a first depth” is meant that each solution is provided at said depth as a discrete solution, and that the solutions get into contact at said depth. For example, each of the acidic solution and the alkali metal silicates solution may be pumped along distinct flow lines to the target injection depth, where they exit said flow lines and mix. Alternatively, the pre-mixed solution can be pumped along a single flow line to the target injection depth, where it exits the flow line.

As the solutions are co-injected, or the mixed solution is injected, at the required depth, the solutions/mixed solution distribute into an underground injection volume. Due to the higher horizontal compared to vertical permeability (horizontal permeability is typically > 10 vertical permeability) the co-injected solutions/mixed solution predominately spread along a lateral direction relative to the vertical. As the solutions distribute laterally, they mix. As a result of the solutions mixing or having been pre-mixed, the acidic solution lowers the local pH such that the silicates in the alkali metal silicate solution start polymerising to form amorphous silica gel. The kinetics of the reaction are such that the mixing solutions continue to spread laterally while the silica-gel progressively forms. In time, the required silica-gel flow diverter forms across the injection volume. Parameters such as the pH characteristics of each solution, the composition of each solution, and the ratio between the injection rates of each solution (in the case of co-injection) or solution ratios (in the case of injection of a mixed solution), can be controlled to fine tune the gelation kinetics of the silica-gel. In particular, said parameters can be tuned to target specific pH values of interest of the underground injection volume.

For a given carbon sequestration reservoir, the acidic solution may be any acidic solution that is capable to reduce the pH of the underground injection volume below a threshold at which silica-gel can form as a stable phase. Suitable examples of acidic solutions in that regard include solutions of hydrochloric acid (HC1), sulphuric acid (H2SO4), and/or nitric acid (HNO3). The solutions may be dilute solutions of said acids.

In some embodiments, the acidic solution is an aqueous acidic solution selected from an aqueous solution of hydrochloric acid, an aqueous solution of hydrochloric acid (HC1), sulphuric acid (H2SO4), and/or nitric acid (HNO3).

The amount of acid in the acidic solution would be sufficient to ensure that, following injection, the underground injection volume has a pH conducive to formation of silica-gel as a stable phase. In some embodiments, the acidic solution is at least a 0.01 M acidic solution, at least a 0.1 M acidic solution, or at least a 1 M acidic solution.

The acidic solution may have a pH that is suitable to ensure that, after injection, the underground injection volume has a pH conducive to formation of silica-gel as a stable phase. For example, the acidic solution may have a pH of about 5 or less, about 4 or less, or about 3 or less.

For a given carbon sequestration reservoir, the alkali metal silicate solution may be any solution leading to precipitation of silica-gel when in contact with the acidic solution. In that regard, the solution would contain an alkali metal silicate that can lead to precipitation of silica-gel under pH conditions at which the silica-gel presents as a stable phase. Suitable pH conditions in the final solution may be a pH below 10. Examples of suitable alkali metal silicates for use in the method of the invention include silicates of sodium (Na), potassium (K), lithium (Li), Cesium (Cs), and Rubidium (Rb).

In some embodiments, the alkali metal solution is a sodium silicate solution.

As used herein, “sodium silicate” identifies chemical compounds of formula Na2 _X Si _y O2y+x or (Na2O) _x -(SiO2)y. Examples of such compounds include sodium metasilicate (Na2SiOs), sodium orthosilicate (Na4SiO4) and sodium pyrosilicate (Na6Si2O?).

For a given temperature, the amount of alkali metal silicate in the solution will be conducive to formation of silica-gel as stable phase under the appropriate pH conditions. In that regard, phase diagrams such as those shown in Figure 2 can provide useful guidance for identifying suitable pH conditions and amount of silicate resulting, for a given temperature, in formation of silica-gel as a stable phase.

In some embodiments, the alkali metal solution contains an amount of silicate of at least about 0.1 M, at least about 0.5 M, at least about 1 M, at least about 2 M, or at least about 5 M, in the form of orthosilicic acid monomer (H4SiO4, represented as “SiO2(aq)” species).

Alternatively, the composition of the alkali metal silicate solution may be characterised in terms of concentration of the alkali metal cations derived from the silicate in solution. Accordingly, in some embodiments the alkali metal silicate solution has a concentration of alkali metal cations of at least about 0.1 M, at least about 0.5 M, at least about 1 M, at least about 2 M, or at least about 5M.

In one aspect the alkali metal silicate solution has a pH sufficiently high to ensure that silicate species present are in a soluble form (e.g. as orthosilicic acid monomers) at room temperature. In some embodiments, the alkali metal silicate solution has a pH of at least about 10.5, for example at least about 11. In some embodiments, the alkali metal silicate solution has a pH of about 12. Co-injection of the acidic solution and the alkali metal silicate solution or injection of a mixed solution may be effected for a duration of time conducive to the formation of an effective silica-gel flow diverter. In some embodiments, the acidic solution and the alkali metal silicate solution / mixed solution are injected for at least about 5 hours, at least about 10 hours, at least about 20 hours, at least about 30 hours, at least about 50 hours, or at least about 100 hours. In some embodiments, the acidic solution and the alkali metal silicate solution / mixed solution are injected for about 20 hours.

The volume ratio of acidic solution and alkali metal silicate solution can be controlled by tuning the ratio of the injection rate of each solution during co-injection, or can be controlled at the mixing stage in the case of injection of the mixed solution. Depending on the specific formulation of each solution, tuning the ratio of the injection rates or volume ratio in the mixed solution can ensure that when the injected solutions are at the required depth, the pH of the resulting mixture is adequate to ensure that silica-gel forms as a stable phase. In some embodiments, the acidic solution and the alkali metal silicate solution are co-injected or are mixed according to a co-injection/mixing ratio of between about 1 to 10 and about 10 to 1, expressed as [flow rate or volume of acidic solution] to [flow rate or volume of alkali metal silicate solution]. For example, co-injection of the acidic solution and the alkali metal solution may be effected by co -injecting the acidic solution at an injection rate of about 2 kg/sec and the alkali metal silicate solution at a co-injection rate of about 12 kg/sec and, for example, mixing of the acidic solution and the alkali metal solution may be effected by mixing the acidic solution and the alkali metal silicate solution at a ratio of about 1 : 6.

In the method of the invention, the acidic solution and the alkali metal silicate solution are co-injected or a mixed solution is injected at a first depth below a first storage volume of the reservoir. Said first depth may be any depth within the reservoir immediately below a suitable storage volume. In some embodiments, said first depth is the bottom of the carbon sequestration reservoir. Said depth can be devised once a suitable storage volume is identified within the reservoir, for example based on the geological characteristics of the reservoir. Typically, said first depth is a depth at which CO2 presents in a supercritical state. In some embodiments, the first depth is a depth of at least about 800 m, at least about 1500 m, or at least about 3500 m.

The acidic solution and the alkali metal silicate solution or mixed solution may be injected at an injection point where the hydrostatic pressure is conducive to effective formation of silica-gel as a stable phase, as well as effective CO2 sequestration. For example, the acidic solution and alkali metal silicate solution or mixed solution injection may be at a depth where the hydrostatic pressure is between about 80 bar to about 350 bar.

Advantageously, the method of the invention may be performed at multiple depths, either sequentially or simultaneously, or a combination thereof. For example, injection of the acidic solution and the alkali metal silicate solution or mixed solution may be performed at a first depth of the reservoir, and subsequently at one or more further depth(s). For instance, the injection of the acidic solution and the alkali metal silicate solution or mixed solution may be performed sequentially at the bottom of the reservoir and at a one or more depths above the bottom of the reservoir. As a result, sequential formation of multiple silica-gel flow diverters at different depths can be effected within the reservoir, each defining a determined storage volume. Alternatively, injection of the acidic solution and the alkali metal silicate solution or mixed solution may be performed simultaneously at one or more further depth(s), one of which may be the bottom of the reservoir. The possibility to provide multiple silica- gel flow diverters at different depths further improves the potential for homogeneous distribution of subsequently injected CO2 within storage volumes of interest.

It may be advantageous to pre-condition the underground volume of interest (i.e. the volume where it is intended to form the flow diverter) by injecting one or more solutions prior to the injection of the acidic solution and the alkali metal silicate solution or mixed solution.

For example, the method of the invention may comprise a step of injecting an acidic solution at the first depth below the reservoir before the injection or co-injection step. Said acidic solution may have the same formulation as the acidic solution used in the co-injection step or mixed solution. The acidic solution permeates mostly laterally relative to the injection site due to a higher horizontal relative to vertical permeability at the injection point. As a result, an initial injection of acidic solution may dissolve carbonate minerals in the form of local cements providing a more homogenous rock layer, which will have an impact on how far the flow diverter can eventually extend laterally. In that regard, a skilled person can readily tune the lateral extension of the flow diverter by controlling the injection parameters (e.g. injection depth, injection duration, injection rate, etc.) of the acidic solution before the coinjection step.

The injection of an acidic solution at the first depth below the reservoir before the coinjection step may be effected for a duration of time conducive to the eventual formation of a silica-gel flow diverter having a desired size. In some embodiments, an acidic solution is injected at the first depth below the reservoir before the injection of solutions or mixed solution for at least about 5 hours, at least about 10 hours, at least about 20 hours, at least about 30 hours, at least about 50 hours, or at least about 70 hours. In some embodiments, an acidic solution is injected at the first depth below the reservoir before the co-injection step for about 30 hours.

The acidic solution may be injected at the first depth below the reservoir before the coinjection or mixed solution injection step at a rate that is adequate for the formation of an effective silica-gel flow diverter having the required size. For example, the acidic solution may be injected at the first depth below the reservoir before the co-injection or mixed solution injection step at an injection rate of at least about 0.1 kg/s, at least about 0.5 kg/s, at least about 1 kg/s, at least about 2 kg/s, at least about 5 kg/s, at least about 25 kg/s, or at least about 50 kg/s.

Efficient formation of the silica-gel flow diverter can also be assisted by pre-conditioning the injection volume before injecting an acid solution at the first depth below the reservoir. Accordingly, in some embodiments the method further comprises injecting water at said first depth of the reservoir prior to injection of an acidic solution. The pre-conditioning using water displaces potentially highly saline and carbonate -rich water providing more defined conditions for the subsequent injection of acid and the co-injection of acid with an alkali metal silicate solution. In some embodiments, the water is groundwater.

In some embodiments, the method further comprises a step of injecting an alkali metal silicate solution at the first depth below the reservoir before the co-injection or mixed solution injection step. Said alkali metal silicate solution may be a solution having the same formulation of that used in the mixed solution or co-injection step. For example, the alkali metal silicate solution injected at the first depth below the reservoir before the co-injection or mixed solution injection step may be an alkali metal silicate solution of the kind described herein.

Injection of an alkali metal silicate solution at the first depth below the reservoir before the co-injection or mixed solution injection step may be effected for a duration of time conducive to the formation of an effective silica-gel flow diverter. In some embodiments, the alkali metal silicate solution is injected at the first depth below the reservoir before the co-injection step for at least about 5 hours, at least about 10 hours, at least about 20 hours, at least about 30 hours, at least about 50 hours, or at least about 70 hours. In some embodiments, the metal silicate solution is injected at the first depth below the reservoir before the co-injection step for about 70 hours.

Injection of an alkali metal silicate solution at the first depth below the reservoir before the co-injection or mixed solution injection step may be effected according to an injection rate that is adequate for the formation of an effective silica-gel flow diverter. For example, the alkali metal silicate solution may be injected at the first depth below the reservoir before the co-injection or mixed solution injection step at an injection rate of at least about 1 kg/sec, at least about 10 kg/ sec, at least about 20 kg/ sec, at least about 50 kg/ sec, or at least about 100 kg/ sec. In some embodiments, the alkali metal silicate solution is injected at a rate of about 12 kg/sec.

A skilled person is well aware of suitable procedures and equipment for effective underground injection of the acidic solution and the alkali metal silicate solution. In some embodiments, the method comprises (i) a step of injecting an acidic solution at the first depth below the reservoir, followed by (ii) a step of injecting an alkali metal silicate solution at the first depth below the reservoir, wherein (i) and (ii) are sequentially performed before the co-injection or mixed solution injection step. In those instances, the acidic solution used in (i) and the alkali metal silicate solution used in (ii) may have the same formulation of the solutions used in the mixture or co-injection step.

In one aspect of the present invention a sequence of four injection phases ensures control over the quality and the extent of the barrier formation, as follows:

1. An acid pre-flush removes minor carbonate cements leading to a more homogeneous injection zone.

2. An alkaline alkali metal silicate solution is injected to avoid direct contact between the acid (phase 1) and the mixed solution made up of acid solution and the alkali metal silicate solution (phase 3) and to allow the alkaline alkali metal silicate solution to be 'pushed' away from the well in phase 4.

3. A mixed solution made up of an acid solution and an alkali metal silicate solution is injected which will lead to gel formation within days.

4. Once gel has formed within the zone invaded by the mixed solution made up of an acid solution and an alkali metal silicate solution (phase 3), CO2 is injected below. The CO2 will migrate upwards towards the lower boundary of the gel, where it will be at least in part diverted and flows laterally away from the injection well. By doing so, it will 'push' the alkaline alkali metal silicate solution (phase 2) to the rim of barrier where further gel is formed. Phase 4 enhances the lateral extend of the flow barrier.

Specific operative conditions of the method of the invention, including the formulation of the acidic solution, the formulation of the alkali metal silicate solution, the ratio of co- injection rate between the solutions, solution ratios adopted in mixed solution, the co- injection time, etc. will be chosen to ensure effective formation of a silica-gel flow diverter that is fit for the purpose of diverting a subterranean CO2 vertical plume into moving laterally.

Typically, the operative conditions of the method will ensure that silica-gel forms as a stable phase after an adequate gelation time.

Conditions leading to the formation of silica-gel as a stable phase have been discussed herein, and are typically governed by the combination of pH in the subterranean volume of interest and amount of silicate. In that regard, phase diagrams such as those shown in Figure 2 provide useful guidance for identifying suitable pH conditions and amount of silicate resulting, for a given temperature, in formation of silica-gel as a stable phase. Said diagrams can inform a skilled person as to adequate formulation of the acidic solution and the alkali metal silicate solution, the ratio of co-injection rate between the solutions, the co-injection time, etc..

In addition, gelation time was also found to be pH-dependent and may therefore be controlled by acting on method operative parameters, including formulation of the acidic solution and the alkali metal silicate solution, the ratio of co-injection rate between the solutions, the co-injection time, etc.. In that regard, tabulates of silica gelation rate relative to the pH are readily available in the literature, and can adequately inform a skilled person as to the choice of those parameters. For instance, for a given temperature silica gelation rate was found to increase by decreasing the concentration of acid and alkali metal silicate in the acidic solution and the alkali metal silicate solution, respectively.

The silica-gel flow diverter formed with the method of the invention is configured such that carbon dioxide injected at a depth below said flow diverter is directed laterally within the reservoir before entering the first storage volume. That is, the characteristics of the flow diverter are such that it is effective to divert a subterranean vertical plume of CO2 by presenting a porosity barrier to the CO2 that contributes to a reduction in permeability to the gas. In that regard, porosity levels of the silica-gel can be controlled by tuning the pH conditions of the subterranean injection volume, and/or by acting on the gelation time prior to CO2 injection. Typically, the lower the pH the lower the porosity of the silica-gel. Without wanting to be limited by theory, it is postulated that lower values of pH result in higher concentration of ionised silanol groups. Said groups mainly present as Si-OH groups that can readily undergo polycondensation into Si-O-Si groups. In addition, longer gelation times prior to CO2 injection typically lead to lower porosity of the silica gel. These principles may offer guidance to the skilled person as to identify a target pH and/or a suitable gelation time prior to CO2 injection for effective formation of silica-gel having porosity characteristics that are adequate to divert CO2.

In some embodiments, the porosity of the silica-gel flow diverter is below about 30%, below about 20%, below about 10%, or below about 5%. In this context, porosity is expressed in terms of % of void per unit of rock volume.

Formation of the silica-gel flow diverter results in a decrease in the volumetric porosity of the formation site. The degree of porosity drop will be dictated by the degree of gelification of the silica-gel. In some embodiments, formation of the silica-gel flow diverter results in a porosity drop of the formation site of at least about 10%, at least about 25%, at least about 50%, at least about 75%, or at least about 90%, relative to the porosity of the formation site prior to formation of the silica-gel flow diverter. By reference to porosity “drop” is meant % reduction of the porosity following precipitation of silica-gel relative to the porosity before gelation. For example, if the pre-injection porosity is 20% and the “drop” in porosity is 50%, it means the post-silica-gel formation porosity is 10%.

The emplacement of the silica-gel reduces the porosity and permeability in the effected volume so that any fluid transport including CO2 is significantly limited under the underground conditions of interest. Generally, the permeability of the silica-gel will progressively decrease as the degree of gelification of the silica-gel increases. In some embodiments, the permeability of the silica-gel is less than about 100, less than about 10, or less than about 1 milli-Darcy. Formation of the silica-gel flow diverter results in a decrease in the permeability of the formation site relative to that of the formation site absent the silica-gel flow diverter. In some embodiments, formation of the silica-gel flow diverter results in a decrease of the permeability of the formation site of at least about 10%, at least about 25%, at least about 50%, at least about 75%, or at least about 90%, relative to the permeability of the formation site prior to formation of the silica-gel flow diverter.

In some aspects of the invention it is desirable for a least some level of CO2 permeation through the flow diverter to take place. In embodiments of the invention where the flow diverter is substantially impermeable to CO2 (Figure la), the CO2 plume is thought to be forced to transition laterally beneath the flow diverter. This can result in a relatively low enhanced storage efficiency and capacity in the vicinity of the storage reservoir directly above the flow diverter, with increased CO2 storage levels above and towards the perimeter of the flow diverter. By tuning the flow diverter formation and CO2 injection to allow at least some CO2 permeability (Figure lb), a proportion of CO2 injected directly below the flow diverter will be able to permeate through the diverter, perhaps at a slower rate than if no flow diverter was present, to allow the storage reservoir space above the flow diverter to attain a higher CO2 storage capacity. For example, it may be desirable to allow up to about 5%, up to about 10%, up to about 15%, up to about 20%, up to about 25%, up to about 30%, up to about 40 % or up to about 50% of CO2 injected below the flow diverter to transition vertically through the flow diverter. For example, this may be achieved by injecting beneath the flow diverter some or all CO2 to be stored before the injected acidic solution and the alkali metal silicate solution / mixed solution has completed gelation. A skilled person can readily determine the appropriate parameters of pH, alkali metal silicate concentration, temperature and time delay before CO2 injection to allow the desired level of permeation of CO2 through the flow diverter (or its precursor acidic solution and the alkali metal silicate solution / mixed solution). For example, where it is desired for the flow diverter to allow permeation of at least some level of CO2 injected below it, the incubation time following the co-injection of the acidic solution and the alkali metal silicate solution or injection of the mixed solution may be as little as about 30 minutes, about one or about two hours, about 3 hours, about 4 hours, about 6 hours, about 8 hours, about 12 hours or up to about a day. The silica-gel flow diverter obtainable with the method of the invention is effective, at least to some extent, in diverting underground flow of CO2 from a vertical direction (as would be the direction of a natural CO2 plume absent the diverter) to a more predominant lateral direction, depending upon the permeability of the diverter. In particular, the flow diverter can cause the flow of CO2 to be distributed out to a greater lateral distance from the injection point. In turn, this increases the area of the vertical base of the CO2 plume, which slows down the vertical flow velocity of the CO2, enhances storage efficiency.

Accordingly, in some embodiments the method of the invention further comprises a step of injecting carbon dioxide at a depth below the flow diverter.

The depth at which CO2 is injected below the silica-gel flow diverter is such that the diverter is effective in deflecting the vertical flow of the resulting CO2 plume. Typically, the CO2 would be injected below and in close proximity to the silica-gel flow diverter. In some embodiments, the CO2 is injected at a depth of less than about 20 metres, less than about 10 metres, less than about 5 metres, or less than about 1 metre below the silica-gel flow diverter.

Due to the depth of the flow diverter, the CO2 present at the injection point is preferably or likely in a supercritical state. Provided it is in a supercritical state at the injection point, there is no particular limitation as to the form in which CO2 is injected from the surface. For example, the CO2 may be injected as a gas from the surface. In other instances, the CO2 may be introduced in the form of a CO2 enriched fluid, for example CO2 enriched water.

The CO2 may be injected after an adequate amount of incubation time following the coinjection of the acidic solution and the alkali metal silicate solution. The duration of said incubation time will be tailored to ensure that the silica-gel has developed porosity and permeability characteristics for effective CO2 flow diversion and/or the level of desired permeation. In some embodiments, the incubation time following the co-injection of the acidic solution and the alkali metal silicate solution or injection of the mixed solution, is at least about 1 day, at least about 2 days, at least about 3 days, at least about 4 days, at least about 5 days, or at least about 10 days.

The CO2 may be injected for any duration of time conducive to the CO2 migrating laterally along the flow diverter and be effectively distributed into the sequestration volume above the diverter. In some embodiments, the CO2 is injected for a period of time of at least about 2 hours, at least about 5 hours, at least about 10 hours, at least about 20 hours, at least about 50 hours, or at least about 100 hours.

In some embodiments, water is injected at the intended CO2 injection point prior to injecting CO2. The injection of water prior to the injection of CO2 can be advantageous in that it provides a buffer zone preventing subsequently injected CO2 from reacting with residual silicates in the injection site, leading to uncontrolled silica-gel precipitation.

In short, in some aspects of the present invention the method of the invention relies on silicate gelation to block high -permeability CO2 pathways. The proposed invention diverts material to a larger volume of the reservoir so that a larger CO2 storage capacity is achieved. That is, the proposed diverter promotes reduced mobility through optimised distribution of the CO2 plume. This is in direct contraposition to conventional mitigation technologies, which rely on CO2 impermeable materials for leak mitigation or leak remediation. However, blocking a CO2 leak differs from diverting a CO2 flow using a flow diverter which distributes CO2 flow through the reservoir so as to promote capillary trapping.

In practice, the proposed method can be advantageous in that it will make CO2 storage safer as a larger proportion of CO2 is no longer stored in form of a mobile plume. Those advantages may make CO2 storage more economical (e.g., by involving lower cost for monitoring) and safer (less mobile CO2) and therefore more acceptable from a risk perspective. The proposed method may also make smaller and unconfined reservoirs (e.g. with no cap rock) attractive for CO2 storage prospective, which would otherwise be dismissed due to the high potential for leakage. The present invention also relates to a subterranean silica-gel flow diverter, said flow diverter extending laterally in a carbon sequestration reservoir at a first depth below a first storage volume of the reservoir, and wherein said flow diverter is configured such that at least some carbon dioxide injected at a depth below said flow diverter is directed laterally within the reservoir before entering the first storage volume.

Said flow diverter may be one that has characteristics described herein, for example one that is obtained in accordance to the method described herein.

As used herein, ambient or room temperature refers to temperatures that may be, for example, between 10°C to 40°C but more typically between 15°C to 30°C. For example, room or ambient temperature may be a temperature between 20°C and 25°C.

Throughout this specification and the claims which follow, unless the context requires otherwise, the word ‘comprise’, and variations such as ‘comprises’ and ‘comprising’, will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

The reference in this specification to any prior publication (or information derived from it), or to any matter which is known, is not, and should not be taken as an acknowledgment or admission or any form of suggestion that that prior publication (or information derived from it) or known matter forms part of the common general knowledge in the field of endeavour to which this specification relates.

Specific embodiments of the invention will now be described with reference to the following non-limiting examples. EXAMPLES

EXAMPLE 1 - Simulation of the formation of an in-depth flow diverter using co-injection of reagents

The following example relates to a simulated formation of a silica-gel flow diverter in a 2D domain. A one-metre thick layer was pre-saturated with groundwater (pH~6) and the rest of the reservoir was saturated with brine (pH~6.5). An acid solution was first injected as an activator solution, followed by injection of a sodium silicate solution. The composition of the solutions is shown in

Table 1 below.

Table 1 - Chemical composition of the solutions used in the computation.

Species Concentration (M)

S .o..dium Ground jwater A . ci.d.i.c so.luti.on Bri ■ne silicate

H ⁺ 1.0E-11 1.0E-6 1.0E-4 1.0E-7

Na ⁺ 1.11 0.001 0.9 0.1

SiO2 _(aq) 1-0 1.OE-11 1.OE-11 1.0E-10

Cl- 0.15 0.001 1.0E-11 0.1

HCO ³ 1.0E-11 1.0E-4 1.1 1.0E-40

Ca ²⁺ 1.0E-11 1.0E-11 1.0E-11 1.0E-40

Mg ²⁺ 1.0E-11 1.0E-11 1.0E-11 1.0E-10

To cool down the one-metre thick layer, the sodium silicate solution was injected at 20°C while the temperature of the other solutions was 60°C, the temperature of the reservoir. Coinjection of the acidic and sodium silicate solutions was defined by adding extra cells to the model and connecting them to the regular grid cells, a feature that is available in the software PetraSim (RockWare®, commercially available from Thunderhead Engineering - https://www.thunderheadeng.com/petrasim).

The methodology used for forming the CO2 flow diverter involved testing different rates of injection and monitoring the main factors that dominate the gelation time. The onset of gelation was expected to occur around day five after co-injection of the acidic solution and the sodium silicate solutions. As such, the injection of the reagents was modelled to occur within 5 days. This was followed by a 5-day no flow (incubation) period to allow the barrier to be formed. Finally, pure CO2 (gas) was injected through the bottom left boundary of the model at a rate of 0.4 kg/s during four days. This step allowed testing the efficiency of the barrier as a flow diverter. Table 2 below shows values used for the operational parameters.

Table 2 - Injection scheme for the simulation for the COtflow diverter reactive transport model

Injection t .ime _ Reagent _ Ra _x te

(hours)

30 Acid 2 kg/s

70 Sodium silicate 1 kg/sm ²

20 _c ^a 2 kg/s and 12 kg/sm

Sodium silicate ²

120 None Incubation

96 CO2 (gas) 0.4 kg/s

In this example, a five-stage injection scheme was used to extend the silica barrier and test its efficiency in diverting the CO2 plume over a 2-week simulation period. Initially, an acid solution was injected into a one-metre thickness layer, which was pre-saturated with groundwater. This was followed by the injection of a sodium silicate solution. This was followed by the co-injection of the acid solution and the sodium silicate solution. After an incubation period of 120 hours, CO2 was injected.

The evolution of porosity, permeability, concentration of silica, and pH at various stages of the procedure is shown in Figure 3. Note, the x-axis has a length of 50 metres and the y-axis a length of 25 metres.

Within the first five days of injection, amorphous silica precipitated 16 metres deep into the highly permeable unit and extended 6 metres more after a 5-day shut-in (Figure 3 A). The outer blue line of the diverter indicates the acid solution (pH~4) injected during the first stage, reacted fairly quickly with the brine of the reservoir (pH~6.5) around the injection point and eased out as it acidified the groundwater (pH~6).

The highest amount of amorphous silica was observed towards the centre of the barrier (red area in Figure 3B), which occurred at values of pH~9. In Figure 3A, the fine yellow rim between the amorphous silica region and the acid front (in dark blue) shows a smooth reaction between these two reagents. The pattern created by the coinjected solutions is seen as a narrow and horizontal lighter blue line (pH~7.11) surrounded by a thicker yellow area at approximate 12 meters deep.

Porosity was reduced by 50% in the area where amorphous silica precipitated the most (light blue area in Figure 3C). The associated drop in permeability was almost 2 orders of magnitude (Figure 3D). Two precursors to silica polymerisation in the form of tetramers (i.e. with four Si centres), were observed to reach their maximum concentration right before the incubation period started. After the 10 days of treatment the underlying concentration of tetramers was elevated near the periphery of the injected reagents (Figures 3E and 3F), where pH values remained close to 10.4 (Figure 3A).

Figure 3G shows the gas saturation (Sg). The barrier contains the gas plume but moreover it deviates its direction, proving its purpose was served. Additionally, some of the gas reacted with the lower side and tip of the diverter (Figure 3H). Despite some reaction between the injected gas and the diverter, the values of porosity and permeability remain in the orders of reduction prior the injection of CO2 (Figures 31 and 3 J) . Finally, when comparing Figure 31 and Figure 3C, the thickness of the area where amorphous silicate precipitated, seemed to have increased during the four days of gas injection.

EXAMPLE 2 - Simulation of barrier formation in a flow-through column experiment

The numerical modelling presented here is based on the original concept by Ito et al. ¹ and the subsequent experimental work by Castaneda-Herrera et al. ^2,3 who considered reactive barrier formation for mitigation and remediation purposes. In this example only the modelling of mitigation cases is discussed, i.e. the domain was initially saturated with a Na- Si solution followed by the injection of an acid. This was carried out to simulate the reaction observed in the laboratory and to be applied at conditions found at the Otway International Test Centre (OITC), Australia ⁴ . The modelling involves simulations of an earlier flow-through column experiment at ambient temperature and pressure conditions using a column with a length of 20 cm and an internal diameter of 2.54 cm. The column was packed with unconsolidated quartz sand yielding a porosity of about 50%. Solutions were injected intermittently at a rate of 1 ml/min. The sequence of injection and no-flow periods was as follows: 20min injection, 8h no-flow; 4min injection, 16h no-flow; 4min injection, 16h no-flow; 7.53h injection. The geometry and parameters used to represent the laboratory set-up are listed in Table 3.

Table 3. Parameters used for the laboratory scale model.

Scenario: Mitigation

Length (cm) 20.0

Diameter (cm) 2.54

Porosity (%) 50

Permeability (D); [m ² ] 10 [9.87E-12]

Pressure (bar) 1.013

Temperature (°C) 20

Injected solution (pH C Ch-enriched

4) * water

Molarity (mol/L) 0.1

Rate (ml/min); [kg/s] 1.0; [1.7E-5]

* Into a column pre- saturated with Na- Si solution (pH=l l).

In flow-through column experiments ^2,3 did not use pure CO2 but CCh-cnrichcd water. Therefore, equation of state (EOS) module EOS1 was used and ID models built to simulate the experiment and the domain divided into 100 grid blocks of 0.2 cm thickness. The volume fractions assigned for quartz and Si02(am) were 0.95 and 0.05, respectively.

A Na-Si solution with a pH of 11 and a viscosity and density similar to water was used to pre-saturate the domain. Once equilibrated in TOUGHREACT ⁵ (commercially available from Lawrence Berkeley National Laboratory - https://tough.lbl.gov/licensing- download/toughreact-licensing-download/), the Na-Si reagent ends up with an initial pH of 10.6, which corresponds to the pH of reagents used by Castaneda-Herrera et al. ² . CO2- enriched waters were simulated by starting with an excess of HCO3- (the master species for CO2 in the TOUGHREACT database), which is converted to CO2 once the water is buffered at a pH of 4. More details of the composition of the two reagents are given in Table 4.

Table 4. Composition of the solutions used to model a flow-through column laboratory experiment: mitigation scenario.

Concentration Concentration

Na ⁺ 1.0 3.0E-4

SiO ₂ (aq) 1.0 1.0E-10

Cl- 0.02 1.0E-10

HCO ₃ - 0.001 0.1

In geochemical databases the molar volume of water- free SiO2(am) is commonly 29 cm ³ /mol and the rate constant for the dissolution and precipitation (k) of SiChiam) is 7.32E-13 mol/m ² /s at 25 °C.

Predicting the rate of dissolution and precipitation and the molar volume of water-containing SiCECgel) is less well constrained. Druhan et al. ^6,7 conducted a modelling study on implementing a hypothetical sealant as a CO2 leakage containment strategy. They suggested the use of a molar volume of 500 cm ³ /mol representing that of a hypothetical gel. Dallemagne ⁸ also followed this approach. Furthermore, Druhan et al. ^6,7 increased the rate constant for the precipitation of the hypothetical sealant (k = 7.3E-08 mol/m ² /s at 25 °C).

In our work, we decided to choose the low rate constant according to Rimstidt and Barnes ⁹ (1980) (Table 3), which is a conservative estimate. The molar volume of the SiO2(gel) was chosen to be 1000 cm ³ /mol, which is equivalent to the 1 molar SiO2(aq) reagent being used in simulations not changing volume upon forming an amorphous silica gel, and thus filling the pore space it occupies once formed.

The modelling of the experimental work allowed testing the chemistry of the solutions as characterized in the laboratory. In TOUGHREACT, the system was charge-balanced using chloride. Once calibrated, these were used throughout the simulations, which provided a basis for evaluating the effects of no-flow periods and molar volume of the SiChCam) phase at the experimental scale.

Simulations were carried out with the objective to test the model performance against the trends observed in permeability and pH behaviour in experimental results by Castaneda- Herrera et al. ³ . Figs. 4 and 5 depict on the y-axis how pH, the abundance of SiChCam), the SiCh saturation index and porosity vary throughout the domain (x-axis), where 0 cm is the inlet or point of injection. The effectiveness of the formation of the reactive barrier seal was measured in relation to the changes in these parameters. A molar volume value of 29 cm ³ /mol for SiChtam) was used in this simulation, which is commonly found in geochemical databases and represents a solid phase SiChCam).

Fig. 4A shows simulation results for a case where the quartz-sand column is first saturated with the alkaline Na-Si solution and CCh-cnrichcd water was subsequently injected for 1 day without any no-flow period. In this case, SiChCam) was expected to occur only at the inlet of the column according to results by Castaneda-Herrera et al. ³ . Once a no-flow period is applied (Fig. 4B) Si02(am) precipitation is expected at the inflow as well as further downstream. The results in Fig. 4B confirm this hypothesis and correspond to a case where CO2-enriched water was injected in four injection / no-flow stages, as follows: 1) 20 minute injection and 8 hours with no flow; 2) and 3) two sets of 4 minute injections with 16 hours of no-flow periods each; 4) a continuous injection was allowed to complete a 2-day simulation as plotted in the green curve, which summarizes the overall behaviour of the simulation.

The first and highest abundance in SiChtam) occurred at the inlet. Three minor peaks were observed downstream at 7, 8 and 10 cm into the domain and were associated with the three no-flow periods (Fig. 4B). The dotted blue line (0.33 h) corresponds to the results at 12 seconds before the completion of the first 20 minutes of injection. The rapid change in the slope observed at 6 cm into the domain, leads to the second maximum and indicates Si02(am) increasingly precipitates as soon as injection ceases. Two subsequent peaks include the amounts of Si02(am) precipitated as a result of the second and third no-flow periods. Despite having the same injection and no-flow time lengths, the third maximum is the highest while the fourth is the lowest.

The results in Fig. 5 were derived using the same conditions as described above for the results shown in Fig. 4 except that the molar volume of SiChCam) was increased from 29 to 1000 cm ³ /mol. A change in pH from alkaline to acidic conditions occurred near the inlet of the column (Fig. 5A). Two zones of SiChCam) precipitation were observed: At the inlet of the column, which formed immediately upon injection and within the zone where the pH changed from alkaline to acidic (Fig. 5B). Only two peaks were observed along the column, at 6.5 cm and 7.5 cm, as compared to the simulation with a lower SiO2(am) molar volume. Furthermore, the simulation using the higher SiO2(am) molar volume yielded a larger amount of precipitated SiO2(am). The results presented henceforth correspond to simulations using the higher SiO2(am) molar volume to represent the silica gel.

The calculated saturation index (Fig. 5C) also corresponded to the observed changes in pH and SiO2(am) precipitation. Values higher than 0 indicate where SiO2(am) is supersaturated, and thus, SiO2(am) precipitation can be expected. After 48 hours the SiO2(am) was no longer supersaturated, therefore most of the Si02(am) present in the column had already formed. A small drop in porosity from 0.5 to 0.46 in the middle of the column (Fig. 5D) was observed.

EXAMPLE 3 - Simulation of barrier formation near the wellbore using sequential injection of reagents

The reservoir chosen is within the Paaratte Formation, which has been used for field-scale CO2 storage demonstration projects ¹⁰ . It is a 7-metre thick, high-porosity interval, which was used for the Otway 2B ¹¹ and Otway 2BX ¹² experiments. The lithology of the Paaratte Formation interval is dominated by quartz (44 wt%) with moderate lithic fragments and a low feldspar content (5 wt%). Clays are found to be pore-filling authigenic kaolinite, chlorite and illite. The dominant cements found in the seal rocks, which frame the reservoir, include dolomite, calcite and some siderite. These bounding cements comprise over 30% grain coating dolomite that has occluded the pore space, reducing the porosity and permeability. This vertically confines the injected fluid near the perforation interval ¹⁰ . Preliminary reaction path modelling suggested minerals such as albite and K-feldspar, among others, have no effect on the precipitation of SiCElam). Therefore, a simplified mineral content was used throughout the model containing quartz, dolomite and SiCElam) only. The kinetic parameters used to describe mineral reaction rates are found in Table 5. Following Druhan et a/., ^6,7 only rate constants applicable for pH-neutral conditions were considered for simplification purposes.

Table 5. Kinetic parameters used in simulations.

Quartz* 87.7 1.023E-14

Dolomite* 52.2 2.95 IE-8

SiO2(am) ^D 60.9 7.32E-13

*Parameters at 25 °C ¹³

^D Parameters at 25 °C ⁹

The relatively homogeneous sandstone reservoir has an average porosity of 28% and average permeability of 2.1 Darcy ¹⁰ and is located between two dolomitic units. The model was hence set up as a 2D radial vertical (x-z) numerical mesh, consisting of two 1 -metre thick seals and one reservoir unit with a total thickness of 7 m and perforated throughout as shown in Fig. 6. The dimensions of the conceptual model used for the near-wellbore cases were similar to those used in previous simulation studies carried out at the OITC ¹¹ . The ECO2N thermophysical fluid property module for FhO-NaCl-CCE ¹⁴ was used.

The numerical mesh comprised 90 layers of constant thickness along the z-axis (0.1 m). Grid spacing in the horizontal direction was implemented, starting with 1.66E-4 m at the well and increased progressively away from it reaching a maximum of 4.9398 m. The model had 9000 cells in total. The top and bottom boundaries were closed and Dirichlet boundary conditions (large block volumes) set to the right side of the model. These large cells simulate the extension of the reservoir beyond the limited dimensions of the basic model. Rock and mineral properties are summarized in Table 6. Table 6. Rock and mineral properties used for near-wellbore reactive-transport models.

Density (kg/m ³ ) 2600 2870

Porosity (%) 28 14

Permeability (k _x ,y) (Darcy) 2.1 0.001

Permeability (kz) (Darcy) 0.21 0.001

Mineral volume percentage (simplified)

Quartz 99.8 73.9

Dolomite 0.1 26.0

SiO2(am) 0.1 0.1

Laboratory results ³ suggested that water-free SiChlam) precipitated at all temperatures studied. Thus, the simulations were conducted in an isothermal system and thermal effects from CO2 injection were neglected and considered as second-order effects. The reservoir temperature and hydrostatic pressure have been reported as 59°C (rounded up to 60°C here) and 14.2 MPa respectively ¹¹ . Initial hydrostatic pressure conditions were set across the domain by running a simulation of 100,000 years using the EOS3.

Both quartz and SiChlam) were allowed to precipitate and dissolve while dolomite was only allowed to dissolve. Following Paterson et al., ⁴ relative permeability models for CO2 gas and liquid water were represented by Corey- and van Genuchten-type curves ¹⁵ , with 0.21 gas and 0.05 liquid residual saturations, respectively. Capillary pressure was derived from the van Genuchten-type model resulting in 26.0 KPa at a water saturation of 0.79.

Simulations at a near-wellbore scale were carried out by means of two different scenarios to evaluate which injection pattern can be seen as a safety measure to avoid gel formation in near-proximity of the well (Table 7). In the first scenario (Case A), a three-stage mitigation run involved pre- saturating the medium with a Na-Si solution followed by the inflow of the SCCO2 source. Finally, the two reactive fluids were pushed away from the well by injecting groundwater without CO2 enrichment. Subsequent scenarios (Cases B-D) included the creation of a buffer zone between the two reagents by firstly injecting the Na-Si solution, then the groundwater and lastly the SCCO2. Two injection periods of groundwater as a buffer were tested (Cases B and C) while SCCO2 was injected for 12 hours. Additionally, the effect of a prolonged injection of SCCO2 (one week) was evaluated (Case D), which was sufficient time to fully flood the domain volume between the wellbore and SiC lgcl) barrier several times (see Fig. 9), and thus, test the stability of the barrier.

Table 7. Scenarios used to run simulations at near-wellbore scale (sequential injection).

No Buffer Case A _XT

„ NaSi - scCO2 - H2O

Zone

Buffer Zone

Case B

Case C

Case D 3 168 co The domain was pre-saturated with Na-Si reagent

* Groundwater

PetraSim, a graphical interface for the TOUGH2 family of simulators was used for this work. TOUGH and TOUGHREACT require an injection rate in kg/s while PetraSim offers a mass flux (kg/(s m ² )) option for the injection rate. The benefit of using the latter option is be that the interface will vary the injection rate for each cell based on the area of the cell. This is necessary when trying to set a uniform boundary condition across cells of varying area and volume. For the near- wellbore simulations, scCO2 and the groundwater were injected at 0.4 kg/(s m ² ).

The compositions of the Na-Si and groundwater solutions used in the near- wellbore models are shown in Table 8. Ca ²⁺ and Mg ²⁺ were added to allow for dolomite dissolution/precipitation in the reactive-transport model. The compositions of the acid solution and brine found in Table 8 were added for the flow diverter case study, as presented below. Table 8. Chemical composition of the solutions for the near-wellbore (Na-Si and groundwater) and for the CO2 flow diverter reactive-transport models (all four solutions).

Na-Si Groundwater Acid Brine

Na ⁺ 1.11 0.001 0.9 0.1

SiO ₂ (a _q ) 1.0 1.0E-11 1.0E-11 1.0E-10

Cl' 0.15 0.001 1.0E-11 0.1

HCO ³ ' 1.0E-11 1.0E-4 1.1 1.0E-40

Ca ²⁺ 1.0E-11 1.0E-11 1.0E-11 1.0E-40

Mg ²⁺ 1.0E-11 1.0E-11 1.0E-11 1.0E-10

Three-stage mitigation scenario without a buffer zone

The model domain was saturated with alkaline Na-Si solution before the injection of scCCh commenced. Fig. 7 shows changes in pH, SiChlam) and porosity after 12 hours of injection of SCCO2 followed by 12 hours of groundwater injection. The wide yellow area towards the left side seen in Fig. 7A represents the groundwater, which had an initial pH of about 6 and becomes more alkaline (~ pH 8.5) during injection and interaction with residual silica reagent. There is a slug of scCCh in-between which should drive the groundwater pH to more acidic conditions. The observed shift in the pH to alkaline conditions could be due to the dissolution of some quartz or Si02(am). However, there is still an increase in the Si02(am) content found in the vicinity of the wellbore (Fig. 7B) after the injection of all three reagents. Fig. 7C shows porosity decreased from 28% to about 22% around the wellbore. The immediate precipitation of Si02(am) near the wellbore upon interaction of CO2 with the Na- Si reagent generated a sharp boundary between the Na-Si reagent, CO2 and water. The sharp boundary may be due to the CO2 being unable to mix into the alkaline Na-Si reagent as it immediately reacted and formed Si02(am). While this scenario is a simplification of the conditions encountered in the field, from the mineralogical, structural and hydrodynamic aspects, it demonstrates the feasibility of isolating the well from the exposure to corrosive fluids. However, the formation of silica gel adjacent to the wellbore is a concern as it may impair injectivity. Therefore, two subsequent scenarios were tested where groundwater was injected prior to the injection of scCCh to create a buffer zone between the two reagents. Three-stage mitigation scenario with a buffer zone

Fig. 8 presents the final distribution of pH, SiC tam) and porosity for two cases involving short-term injection of water prior to the injection of scCCh (Table 7): Case B had 3 hours of water injection, and Case C had 6 hours of water injection prior to the injection of scCCh for 12 hours. Overall, the results show similar changes in various properties (pH, Si02(am) precipitation) as the CO2 interacts with the Na-Si reagent in the reservoir. Importantly, the integrity near the wellbore is preserved. Additionally, the results of Case B indicate that only a small volume of groundwater is needed as a buffer zone and that this allows the barrier to form at a specific location away from the wellbore within the reservoir. This behaviour could be associated with the groundwater acting as a buffer zone between the acidic CO2 and the alkaline Na-Si zones. The pH behaved as expected where it decreased as the CO2 was injected (close to that of a CCh-cnrichcd water, pH ~ 4.5) and increased at the interface with the Na-Si reagent. As a result, the abundance of Si02(am) precipitate increased from 0.0719 to 0.18 mol/m ³ with an associated drop in porosity (about 70%).

Increase in lateral extent of silica barrier by CO 2 migration

Case B was subsequently modified by injecting groundwater for 3 hours followed by the injection of scCCh for 1 week instead of 12 hours. The results of this simulation are presented in Fig. 9. The plume reached a radius of approximately 17 m. Fig. 9A shows the pH dropped from 11 (initial pH of Na-Si reagent in the reservoir) to about 6 at the plume front. Near the wellbore, pH dropped as low as 4.5. The change in pH is enough to increase the abundance of Si02(am) from 0.07 to about 0.16 mol/m ³ , as seen in Fig. 9B. Porosity changes are presented in Fig. 9C and indicate a wide area of reaction. The stepwise change in the predicted pH, Si02(am) and porosity at the interface between the plume front and the preinjection reservoir (Fig. 9) is considered to be an artefact relating to the very steep gradients between neighbouring grid cells.

Additional tests were conducted to evaluate how the barrier would age after the well was shut down. Results after 3 and 6 months are presented in Fig. 10 and Fig. 11, respectively. The lateral extent of the plume in Fig. 10 indicates that CO2 continued migrating and interacting with the reagent, mostly towards the top seal (compared with Fig. 9). However, the extent of the plume after 6 months (Fig. 11), was not much different than after 3 months. This flow reduction could indicate that further SiO2(am) precipitation may have impeded CO2 migration.

References

1. Ito, T., Xu, T., Tanaka, H., Taniuchi, Y., Okamoto, A., 2014. Possibility to remedy CO2 leakage from geological reservoir using CO2 reactive grout. International Journal of Greenhouse Gas Control 20, 310-323.

2. Castaneda-Herrera, C.A., Black, J.R., Stevens, G.W., Haese, R.R., 2017. Preliminary experiments for a chemical reactive barrier as a leakage mitigation technology. Energy Procedia 114, 4140-4146.

3. Castaneda-Herrera, C.A., Black, J.R., Llanos, E.M., Stevens, G.W., Haese, R.R., 2018a. Formation of an Amorphous silica Gel Barrier under CO2 Storage Conditions. International Journal of Greenhouse Gas Control 78, 27-36.

4. Paterson, L., Boreham, C., Bunch, M., Ennis-King, J., Freifeld, B., Haese, R. R., Jenkins, C., Raab, M., Singh, R., Stalker, L., 2011. The CO2CRC Otway stage 2B residual saturation and dissolution test: test concept, implementation and data collected. CO2CRC Report No.: RPT11-3158, p. 72.

5. Xu, T., Spycher, N., Sonnenthal, E., Zhang, G., Zheng, L., Pruess, K., 2011. TOUGHREACT Version 2.0: A simulator for subsurface reactive transport under nonisothermal multiphase flow conditions. Computers and Geosciences 37: 763-774.

6. Druhan, J.L., Vialle, S., Maher, K., Benson, S., 2014. A reactive transport model for geochemical mitigation of CO2 leaking into a confined aquifer. Energy Procedia 63, 4620-4629.

7. Druhan, J.L., Vialle, S., Maher, K., Benson, S., 2015. Numerical simulation of reactive barrier emplacement to control CO2 migration. Carbon Dioxide Capture for Storage in Deep Geological Formations - Results from the CO2 Capture Project, Volume 4, CCS Technology Development and Demonstration Results (2009-2014), Chapter 45, 825-845.

8. Dallemagne, T.F.M., 2017. Numerical simulation of near wellbore permeability reduction using colloidal silica gel. Master’s Thesis, Utrecth University. Rimstidt, J.D., Barnes, H. L., 1980. The kinetics of silica-water reactions. Geochimica et Cosmochimica Acta 44, 1683-1699. Paterson, L., Boreham, C., Bunch, M., Dance, T., Ennis-King, J., Freifeld, B., Haese, R.R., Jenkins, C., LaForce, T., Raab, M., Singh, R., Stalker, E., Zhang, Y., 2013. Overview of the CO2CRC Otway residual saturation and dissolution test. Energy Procedia, 37, pp. 6140-6148. Haese, R.R., LaForce, T., Boreham, C., Ennis-King, J., Freifeld, B.M., Paterson, L., Schacht, U., 2013. Determining residual CO2 saturation through a dissolution test - Results from the CO2CRC Otway Project. Energy Procedia, 37, pp. 5379-5386. Paterson, L., Ennis-King, J., LaForce, T., Gilfillan, S., Semo, S., Johnson, G., Boreham, C., Dance, T., Freifeld, B., Singh, R., 2015. CO2CRC Otway 2B Extension Phase 2. CO2CRC Report No. RPT15-5429. Palandri, J.L., Kharaka Y.K., 2004. A compilation of rate parameters of water-mineral interaction kinetics for application to geochemical modelling. U.S. Geological Survey Report No. 1068. Pruess, K., Spycher, N., 2007. ECO2N - A fluid property module for the TOUGH2 code for studies of CO2 storage in saline aquifers. Energy Conversion and Management, 48(6) 1761-1767. van Genuchten, M.T., 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society of America Journal, Volume 44, 892-898. Stavland et al., 201 la, In-depth water diversion using sodium silicate - preparation for single well field pilot on Snorre. Proceedings of the European Symposium On Improved Oil Recovery, Cambridge, UK, 12 - 14 April. Stavland et al., 2011b, In depth water diversion using sodium silicate on Snorre - factors controlling in-depth placement. Proceedings of the SPE-143836 European Formation Damage Conference, Noordwijk, The Netherlands, 7 - 10 June. Stavland et al., 2015, Planning for in-depth diversion using sodium silicate. Tech. Rep. IRIS 2015/157, International Research Institute of Stavanger.