Hydrogen from Depleted/Depleting Hydrocarbon Reservoirs: A Reservoir Engineering Perspective

Abstract

In today’s industry, H₂ is mostly produced from fossil fuels such as natural gas (NG), oil, and coal through various processes. However, all these processes produce both carbon dioxide (CO₂) as well as H₂, making them questionable in terms of climate change mitigation efforts. In addition to efforts to increase the conversion efficiency of green H₂ technologies, work is also underway to make H₂ production from fossil fuels more environmentally friendly by reducing/avoiding CO₂ emissions. In this framework, these technologies are combined with geologic carbon storage. In a further step, the use of depleted hydrocarbon reservoirs for in situ H₂ production is being investigated, with the co-generated CO₂ remaining permanently in the reservoir. The objective of this paper is to provide a brief overview of the technologies that can be used to produce H₂ from depleted and depleting hydrocarbon reservoirs (DHRs) in various ways. We evaluate the required processes from a reservoir engineering perspective, highlighting their potential for H₂ generation and their technology readiness level (TRL) for applications. We also investigate the possibility of permanently storing the co-produced CO₂ in the reservoir as a means of mitigating emissions. In addition, we provide a preliminary cost analysis to compare these methods with conventional hydrogen production techniques, as well as an assessment of operational risks and associated cost estimates.

1. Introduction

Renewable hydrogen plays a critical role in energy transition as it can replace fossil fuels (such as gas, coal, and oil) in industries and transportation that are challenging to decarbonize [1]. The EU report has set targets for both domestic hydrogen (H₂) production and imports that align with the zero-emission goal of the International Energy Agency (IEA) by the end of 2050. The plan will be enhanced by establishing H₂ infrastructure for the production, transport, and trade of 20 million tons of H₂ by 2030 [1]. In the USA and other industrialized nations, the importance of H₂ in the energy sector is set to grow as these countries establish ambitious goals to increase its usage. For example, the Department of Energy (DOE) in the USA announced the “Hydrogen Shot, 1-1-1” initiative, aiming to reduce the cost of clean H₂ to 1 USD per 1 kg of H₂ within a decade. This target represents an 80% reduction from its current estimated average cost of 5 USD per kg [2].

Hydrogen is the most abundant element in the universe and has the maximum energy content on a mass basis compared to any other known fuel. Using H₂ as an energy carrier is environmentally friendly as it does not produce any harmful emissions. The only byproducts generated are heat and water [3,4,5]. In today’s industry, H₂ is predominantly produced from fossil fuels such as natural gas (48%), oil (30%, mainly consumed in refineries), and coal (18%). Several processes are used to generate H₂, including thermal cracking, catalytic cracking, and the most common one, steam methane reforming (SMR). All of these processes, however, produce carbon dioxide (CO₂) alongside H_2, raising concerns about their compatibility with climate protection efforts. A modern SMR plant, for instance, produces approximately 9 kg CO₂ per kg of H₂ [5,6]. The proportion of the CO₂-neutral H₂ production from renewable resources (green H₂), such as in water formed by electrolysis or bioprocesses, is currently small (2% of overall H₂ production in 2020). The current price of H₂ with a fossil origin, including the removal of co-produced CO₂, is approximately 1–3 USD/kg (SMR and other conventional technologies) [5,7]; the cost of green H₂ generation technologies is as high as 10 to 20 USD [5,8].

In addition to enhancing the conversion efficiency of green H₂ production methods, researchers are also working on making fossil fuel-based H₂ production more environmentally friendly by reducing CO₂ emissions. Within this framework, these technologies are integrated with geologic carbon storage (GCS) to safely store the captured CO₂ in specific geological formations. GCS transforms grey H₂ production, where CO₂ is released into the atmosphere, into blue H₂, which is considered climate-neutral. In a subsequent step, researchers are exploring the possibility of generating H₂ in situ within depleted or depleting hydrocarbon reservoirs (DHRs) while permanently storing the co-generated CO₂ in the reservoirs to eliminate emissions.

The objective of this paper is to provide a brief overview of the technologies that can be used to extract H₂ from DHRs in a variety of ways. We assess the claimed processes from a reservoir engineering perspective, emphasizing their H₂ generation potential and technology readiness level (TRL) for applications. We also investigate whether it is possible to leave the co-produced CO₂ permanently in the reservoir in order to reduce emissions. Additionally, we provide a preliminary cost analysis to compare these methods with traditional H₂ generation techniques, along with an evaluation of operational risks and their associated cost assessments.

2. Depleted/Depleting Hydrocarbon Reservoirs (DHRs) for H₂ Generation

Oil and gas reservoirs typically become depleted within a few decades, with the exact timeframe depending on factors like a reservoir’s size, its characteristics, and the rate at which it is being depleted. Natural depletion refers to the process of producing all of the recoverable hydrocarbons using the energy of the reservoir. However, due to thermodynamic and petrophysical limitations, a portion of the hydrocarbons initially present in the reservoir cannot be produced and remain there; in depleted oil reservoirs (DORs), 50% to 90% of the original oil in place (OOIP) remains unproduced. In contrast, primarily due to the higher compressibility of the gas, depleted gas reservoirs (DGRs) have a significantly lower percentage of unproduced resources, ranging from 10% to 40%. Enhanced oil recovery (EOR) methods are employed to extract a portion (5% to 20%) of the remaining oil in the reservoir by injecting chemicals and/or energy from the surface, thereby improving production efficiency. In fact, there would still be a reasonable amount of oil left in the reservoir. In gas reservoirs, the methods used to recover the remaining gas after natural depletion are often considered not economically viable because of the additional costs of separating the natural gas (NG) from the injected gas.

Within the context of optimizing and decarbonizing the remaining energy resources, the further utilization of depleted hydrocarbon reservoirs (DHRs) is a topic frequently discussed in the literature. A common proposal is to use the “emptied” reservoir volumes for CO₂ storage in an effort to reduce the carbon footprint of various industries. This type of geologic carbon storage (GCS) has evolved into its own industry. The technology has been discussed and developed for many decades, gradually reaching an industrially applicable level. Reportedly, a CO₂ storage capacity of approximately 75 million metric tons per year (Mt/year) will be achieved by 2024 in current and under-construction GCS projects, including the reuse of DHRs [9]. The global geological CO₂ storage resources in DHRs are estimated at up to 1000 gigatons (Gt) of CO₂ (Alkan et al. 2023) [10].

The related literature discusses alternative solutions for optimizing the use of DHRs in energy transition efforts. One proposed solution appears to offer a twofold benefit: producing H₂ in situ using the remaining hydrocarbons and leaving the co-generated CO₂ in the reservoir.

Table 1. Overview of the proposed technologies using DHRs (DORs and DGRs) for generating H₂ in situ.

Method	Process Proposed	Selected References
Microbial H2 generation (DORs)	Dark fermentation (anaerobic) of sugar. This is a derivative of microbial EOR (MEOR) to improve oil production.	[11,12]
Catalytic methane conversion (DGRs)	In situ steam methane reforming (SMR), water–gas shift (WGS) reaction, and methane cracking (MC) requiring heat (350–450 °C with steam flooding); for catalytic MC, Ni-salt-based solutions are used.	[13,14,15]
H2 through in situ combustion (DORs)	Converting in situ hydrocarbons of DORs into H2 using reforming techniques. Besides hydrocarbon, the other reactants can involve steam (steam reforming) or oxygen (partial oxidation, with up to 1000 °C using in situ combustion) or both (auto-thermal).	[15,16,17]

summarizes the three technologies discussed in the literature that are applied in this context, along with relevant references for each. The principles of these technologies are similar to those used for H₂ production from hydrocarbons in surface installations. A common feature among all these technologies is the injection of an energy source from the surface, allowing sufficient time for in situ reactions to produce gases including H₂ and CO₂. The lighter H₂ is then separated from heavier components such as CO₂ either in situ or through separation techniques at the surface or downhole. The H₂ is then transferred to the place of use.

The schematic overview of generating H₂ from DHRs is depicted in Figure 1.

A substrate to feed the microorganisms in the reservoir is injected from the surface or the reservoir is heated with injected energy to initiate a steam flooding or in situ combustion process. The injection triggers the proposed reactions in the reservoir, resulting in H₂ and the production of accompanying gases, like CO and CO₂. It is claimed that lighter H₂ separates rapidly from the other gases and moves upwards, accumulating at the top of the reservoir. CO₂ or other gaseous components like methane (CH₄) can dissolve in water and oil. In a gas reservoir, CO₂ tends to move downwards as its density is higher than that of NG. It is certain that more expensive horizontal wells can increase the performance of the processes as the energy injected is better spread throughout the reservoir. It is argued that the membrane separators can be used to capture H₂ from the gas streams already downhole [18,19]. For higher amounts of H₂ and/or in a case where such separation is not possible, the separation of the H₂ and other gases can be performed using capture units installed at the surface, which, however, involves higher investment costs.

The related literature lacks an in-depth engineering analysis of these processes. Are the technologies mature enough and are the processes technically and economically viable? What are the main drawbacks and risks of H₂ generation in the reservoir? Is in situ H₂ generation, production, and separation at the surface competitive with conventional H₂ generation methods? In the following sections, we describe these technologies and assess their technical and economic feasibility.

3. Microbial H₂–Bio-H₂ Generation in DORs

3.1. Concept

The generation of H₂ from biomass and/or organic wastes through biological and chemical processes is a well-defined, commercially applied process [20,21,22]. The energy required for biochemical reactions that generate H₂ can be provided by solar power or by converting fluid carbon sources, or by a combination of both. The conversion of carbon sources involves microbial metabolism, and one of the related technologies is called dark fermentation (DF). DF refers to the degradation of organic substrates by anaerobic bacteria in an environment without a light source or oxygen to produce bio-H₂. The DF is industrially tested in up-scaled bioreactors, creating the required thermodynamic and physiological environments to generate optimum amounts of bio-H₂ [22,23,24]. Numerous studies have investigated the microorganisms with their origins and capacities in terms of DF under various conditions [25,26,27,28,29,30]. Clostridiaceae, Methylotrophs, Enterobactericeae, and Thermoanaerobacterium are anaerobic or facultative genera of microorganisms commonly utilized for their superior H₂ production capacity [22,28,30,31]. DF bacteria can decompose a variety of substrates (formic acid, lactic acid, cellulose disaccharide, sulfide, etc.) into H₂ under the action of nitrogenase or hydroenzyme. Raw materials such as food and agricultural wastes are said to trigger chemical reactions that produce H₂ at relatively competitive prices. However, the H₂ yield and production rate are low, presenting a significant challenge for large-scale industrial applications. Various publications discuss economic indicators for small DF units, and based on the provided data, the cost of H₂ production shows a wide range from 2.5 to 34 USD₂₀₂₃ per kg of H₂, with daily production rates of up to 10 tons [21,32,33]. This cost range is considerably higher than the actual costs of any conventional H₂ production. Moreover, despite its low global warming potential, DF exhibits very poor energy and exergy efficiencies [33]. The main technical barriers in industrial applications under discussion are the development and operation of functional bioreactors and the thermodynamic limitations on the H₂ yield by microbial fermentation. The availability of raw materials and their costs are cited as the primary economic barriers to the economic potential of large-scale DF applications.

The energy produced from glucose (or sucrose) is often cited as one of the primary and most effective examples of DF reactions:

C₆H₁₂O₆ + 2H₂O → 2CH₃COOH + 4H₂ + 2CO₂

(1)

This reaction involves fermentative microorganisms and produces two useful products: acetic acid (CH₃COOH) and acetate, as well as H₂, after a certain incubation period. Brudhoo and Mohee have conducted an in-depth and critical review of various factors involved in the DF process [34,35]. These factors include H₂ consumers, bacteria in different media, and inhibitory factors. The authors also provided a summary of the principal biochemical reactions that lead to bio-H₂ generation, and they presented the H₂ yield resulting from this mechanism, as reported in various studies [20,25,27,28,31,36]. Table 2 presents the results of selected experimental works including the H₂ yield per type of carbohydrate used, the microorganisms involved, and the experimental conditions.

The use of depleted and even depleting hydrocarbon reservoirs for bio-H₂ production is a derivative of the DF efforts discussed earlier at both the laboratory and industrial scales. It is well known that hydrocarbon reservoirs host various types of microorganisms as long as the thermophysical conditions allow this. Based on available data and considering that molecular oxygen (O₂) is rapidly removed by microbial and/or chemical processes, it is suggested that only strict anaerobic bacteria can be considered truly indigenous to oil reservoirs [37]. This suggestion is supported by many studies that confirm the widespread presence of anaerobic biospheres in oil reservoirs [38,39]. The concept of generating bio-H₂ involves using the reservoir as a bioreactor, with various metabolic pathways including fermentation, anaerobic hydrocarbon degradation, acetogenesis, methanogenesis, and other anaerobic respiration processes such as nitrate or iron reduction, as well as the reduction of oxidized sulfur species, in addition to DF.

Although the idea is relatively new, the production of H₂ in a reservoir due to fermentative microorganism reactions is known from the literature related to microbial enhanced oil recovery (MEOR) [40,41]. Recent studies have reported experiments mimicking the process in the reservoir, suggesting field trials [11,12]. The H₂ yields obtained in these studies are also provided in Table 2. As shown in the table, using oil as a standalone carbon source is not promising in terms of H₂ generation. However, in a study, when 1 g/L of glucose and 25 mg/L of a surfactant were added to improve the interaction of oil with aqueous phases (thus making hydrocarbons more accessible to microorganisms in the aqueous phase), H₂ generation was enhanced compared to the case in which oil was the only substrate [11].

The following results can be derived from the table and related references.

Many different microorganisms are capable of initiating DF. In reservoirs where the temperature is less than 80 °C, it is common to find one or a community of such microorganisms. Their physiology and mechanisms of action are well studied in the related industry. However, the microbiology of reservoirs is field-specific and should be analyzed to determine their potential for H₂ generation.

H₂/sucrose yields slightly higher than 4 (in mol basis), as reported in a few references, are due to the use of other carbon sources and additives such as yeast. However, the addition or use of low-quality carbohydrates (such as molasses and starch) decrease the H₂ yield.

Using in situ hydrocarbons as standalone carbon sources is not effective for generating bio-H₂. However, adding low concentrations of glucose and surfactants in the aqueous phase can increase the yield. It should be noted that the yield is still lower than that obtained with the maximum possible carbohydrate concentration.

Aquifers containing beneficial bacteria (e.g., Enterobactericeae and Thermoanaerobacteriumas discussed previously) and thermophysical conditions can also be used for bio-H₂ generation with comparable yields.

The high pressure in the reservoir reduces microbiological activity, thus decreasing H₂ generation.

Unfortunately, there is not enough conclusive dedicated dynamic data (corefloods) supporting the H₂ generation in a DOR; however, data from MEOR studies confirm H₂ generation in porous media, although not as much as the results of batch experiments as indicated in Table 2 [42].

Table 2. Results of H₂ generation through DF at laboratory scale from selected references.

Carbon Source	Experiment	Max H2	Microorganisms	Ref.
Used		Yield 1
Glucose	Bottle, atm	4.08	Escherichia coli	[43]
Glucose	Bioreactor, atm, 35 °C	1.65	Clostridium species	[25]
Sucrose	Bioreactor, atm, 35 °C	4.52	Clostridium species	[25]
Sucrose	Bottle, atm	2.98	Clostridium butyricum	[28]
Glucose	Bioreactor, atm, 32 °C	3.80	Clostridium pasteurianum	[31]
Glucose	Bottle, atm, 70 °C	1.10	Thermotoga strains	[11]
Crude oil	Bottle, atm, 70 °C	0.10	Thermotoga strains	[11]
Glucose + crude oil	Bottle, atm	1.65	Thermotoga strains	[11]
Glucose	Bottle, atm, 37 °C	1.12	Reservoir community 2	[12]
Glucose	Bottle, 16 bar, 37 °C	0.70	Reservoir community 2	[12]

¹ In mol/mol carbohydrate. ² Community containing Halanaerobium.

Table 2. Results of H₂ generation through DF at laboratory scale from selected references.

Carbon Source	Experiment	Max H2	Microorganisms	Ref.
Used		Yield 1
Glucose	Bottle, atm	4.08	Escherichia coli	[43]
Glucose	Bioreactor, atm, 35 °C	1.65	Clostridium species	[25]
Sucrose	Bioreactor, atm, 35 °C	4.52	Clostridium species	[25]
Sucrose	Bottle, atm	2.98	Clostridium butyricum	[28]
Glucose	Bioreactor, atm, 32 °C	3.80	Clostridium pasteurianum	[31]
Glucose	Bottle, atm, 70 °C	1.10	Thermotoga strains	[11]
Crude oil	Bottle, atm, 70 °C	0.10	Thermotoga strains	[11]
Glucose + crude oil	Bottle, atm	1.65	Thermotoga strains	[11]
Glucose	Bottle, atm, 37 °C	1.12	Reservoir community 2	[12]
Glucose	Bottle, 16 bar, 37 °C	0.70	Reservoir community 2	[12]

¹ In mol/mol carbohydrate. ² Community containing Halanaerobium.

3.2. Reservoir Engineering Aspects

With reference to both existing studies as well as studies on similar processes (EOR, mainly MEOR), the following description of a potential bio-H₂ process in a DOR can be made.

3.2.1. Mode of Application

A solution of carbohydrate (which may be supplemented with other feeding chemicals) is injected into the selected DOR with various degree of depletion. A bioreactor is expected to form in the reservoir at a distance from the well, which depends on the incubation time, injection rate, reservoir geometry, and petrophysics. Based on laboratory studies, which have defined incubation times of less than 10 days, and the injection rates that are mostly applied, the distance of the bioreactor from the injector cannot exceed 10–20 m. During the process, H₂ is generated, along with other reaction products such as organic acids, CO₂, and H₂S. The H₂ can be separated from the gas stream downhole or at the surface. The CO₂, which is more soluble in water and oil compared to H₂, as shown in Figure 2, tends to remain in the reservoir if not produced alongside oil and water [44,45]. The process is a derivative of MEOR, which means that a portion of the remaining oil after the primary and secondary recovery phases can be mobilized. This can increase economic figures, but it raises questions about separating produced gases from the oil at the surface. Additionally, there are potential negative environmental impacts of co-producing a carbon-intensive fluid like oil, especially if it is utilized at a later stage. From a technical standpoint, the process is similar to EOR analogies used in the oil industry, which means that it has a TRL of 7–8.

3.2.2. Economics

Because of potentially high adverse effects (like plugging due to biomass excess [42]), the carbohydrate concentration to be injected is limited to ca. 10 mg/L or approximately 0.05 moles/L. Independently, the max. H₂ yield of DF is 4 mole H₂/mole glucose. This means 180 g of glucose generates a maximum of 8 g H₂. Assuming the price of glucose as 0.7 USD₂₀₂₃/kg at the international market (which is currently cheaper compared to historical levels), 1 kg of H₂ generated would cost 15.75 USD_2023, based on the carbohydrate cost. To obtain the theoretical yield, additives like yeast, which costs 6 USD₂₀₂₃/kg, are often included to the reactant in the experiments. This range of costs for the substrate, which has to be injected in a reservoir, is comparable with the H₂ production costs in surface DF processes if similar high-grade feedstocks like glucose are used [21,33]. Assuming the theoretical yield is attainable under reservoir conditions, and also considering other operational and relatively minor CAPEX (capital expenditure) costs, with actual technology, the unit cost from microbial in situ bio-H₂ generation is likely to be higher than 25 USD₂₀₂₃/kg. Injectant solutions may contain higher (e.g., sucrose) or lower (e.g., molasse) potent carbohydrates, which are associated with high yields/high costs vs. low yields/low costs per unit substrate, respectively. However, the ratios of yields to substrate costs are approximately equal for both cases, which leads to similar H₂ production prices irrespective of the carbohydrate quality.

3.2.3. Associated Risks

An in situ bio-H₂ operation can be impacted by various risks that can violate both technical and economic criteria. The risks that can be relevant to the application of MEOR are of primary importance for potential field applications of bio-H₂ [49]. The presence of sulfate-reducing bacteria (SRB) can lead to bio-souring; H₂S production is one of the main show-stoppers in all oil and gas operations, making it a particularly concerning risk [50].

The inhibitory factors on DF are discussed in [34]. This paper provides an in-depth and critical review of various factors that can impact the DF process. These include H₂ consumers and lactic acid bacteria in mixed microflora, as well as slight and heavy metal ions and furan derivatives that can negatively influence H₂ production. These factors can create operational risks that need to be considered when implementing the process. The thermodynamic conditions, namely pressure and temperature, also influence the effectiveness of the DF. It is indicated that higher pressure reduces the production of H₂ whereas the reservoir temperature should be less than the maximum temperature at which the fermentative bacteria can survive [12]. Note that the costs of associated risks for the mitigation and prevention of H₂S souring and fouling, as well as the separation, transport, and elimination of produced CO₂, were not included in the preliminary cost estimation provided above.

The key questions in engineering bio-H₂ in DORs are how and why microbial in situ H₂ generation could be competitive with DF bioreactors at surface plants. When considering operational costs alone and based on theoretical and laboratory yields, the estimated cost of bio-H₂ production at a commercial scale is higher than that of any other H₂ production process. Additionally, there is a potential conflict of interest in using high-quality carbohydrates such as glucose and sucrose for H₂ generation instead of using them as food for humans and livestock.

4. Thermal H₂ Generation from Hydrocarbons: Theoretical Basis

The concept of in situ H₂ generation in DHRs is based on the chemical transformations of hydrocarbons under the thermal effect. Currently, several mature technologies produce H₂ from fossil fuels at the surface, with hydrocarbon reforming and pyrolysis being the most utilized for the actual H₂ demand. The chemical processes used are well described and studied and can be found elsewhere [8,51,52]. As the same reactions are valid for the production of H₂ from DHRs, they are briefly elaborated in the following sections.

4.1. Steam Reforming

This is the reaction of a mixture of steam and hydrocarbons to produce H₂ and carbon oxides:

C_nH_m + H₂O ↔ nCO + (m/2 + n)H₂ ΔH > 0

(2)

The reaction is highly endothermic. The most widely used hydrocarbon reformation process is SMR from NG or light hydrocarbons. To overcome or reduce some of the shortcomings of steam reforming, such as coke formation, high-performance catalysts are generally required. These catalysts are commonly based on nickel, although precious metals are also active but too expensive for commercial application. The nickel catalysts are supported on ceramic oxides or oxides stabilized by hydraulic cement. The steam reforming reaction is further enhanced by the water–gas shift (WGS) reaction, in which carbon monoxide and water are converted to CO₂ and H₂:

CO + H₂O ↔ CO₂ + H₂ ΔH = −41 kJ/mol

(3)

4.2. Partial Oxidation

Partial oxidation is mostly claimed to be the most appropriate technology for obtaining H₂ from heavy fuel oil and coal. Partial oxidation is an exothermic process used in converting hydrocarbon fuels into a mixture of H₂, CO, and other partially oxidized substances. One of the advantages of this process is that reactions with O₂ are highly exothermic, without the necessity of any external energy source. The product distribution of partial oxidation reactions depends upon the carbon/oxygen (C/O) ratio and is constrained by high reaction temperatures (>1000 °C). Depending on the sulfur content in the feedstock, the partial oxidation process is either operated without catalyst as a pure thermal partial oxidation process (TPOX) at a temperature >1200 °C if the sulfur content is high or at a lower temperature, between 800 °C and 900 °C, using a solid catalyst if the sulfur content is considerably low.

The catalytic partial oxidation at lower temperature is more energy-efficient and globally described by the following reaction:

C_nH_m + ½n O₂ → nCO + ½m H₂ ΔH < 0

(4)

Thermodynamically, in this process, H₂ and CO are the most abundant products above 550 °C, with CO being a coke precursor that can be removed by its oxidation toward CO₂ or by the WGS reaction (Equation (3)) increasing H₂ production.

4.3. Autothermal Reforming

Autothermal reforming includes the exothermic partial oxidation with O₂, which provides the energy needed for the endothermic steam reforming reactions. In essence, both steam and oxygen are introduced into the reformer, leading to the reforming and oxidation reactions simultaneously occurring to obtain a thermodynamically neutral reaction. As in steam reforming or partial oxidation, catalyst selection plays a crucial role in the overall performance, with nickel-based catalysts being most commonly used because of their effectiveness and low cost. Given the high thermal efficiency of this process, it requires lower energy than steam reforming or partial oxidation. A higher H₂ yield is produced in autothermal reforming than in partial oxidation; the yield is, however, lower in the steam reforming process. In thermal cracking or thermolysis, coke (solid hydrocarbon) and H₂ are also produced by the action of the heat excess.

4.4. Pyrolysis

Pyrolysis (thermal cracking) is a thermal decomposition process occurring in the absence of oxygen, which converts light hydrocarbons into elemental carbon and H₂:

C_nH_m ↔ nC + m/2H₂ ΔH > 0

(5)

If the solid carbon (coke) is generated as depicted in Equation (5), it can be considered sequestrated in the reservoir. At the same time, side reactions also occur that consume generated H₂. These reactions mostly include methane-forming reactions like reverse CO₂ reforming:

2CO + 2H₂ ↔ CH₄ + CO₂ ΔH = 59.0 kcal/mol

(6)

Methane pyrolysis is widely reported in the literature as not producing CO₂ because all the carbon is in solid form. In practice, NG is used as a source for industrial processes instead of pure CH₄; hence, other compounds such as CO₂, H₂O, and higher hydrocarbons are involved in addition to CH₄. These compounds also react during the pyrolysis process, which strongly affects product selectivity and conversion. Various catalysts are known for this process, ranging from metallic catalysts to carbonaceous catalysts.

The reactions (2) through (5) generate H₂. The H₂ yield decreases with an increase in hydrocarbon molecular weight, i.e., when the ratio of hydrogen and carbon (H/C) in chemical compounds decreases. Thus, yield decreases in the following sequence: methane -> lighter oil -> heavier oil -> coal. From a thermodynamic standpoint, depending on the conditions, reactions (5) and (6) can both consume H₂ to generate CH₄. This implies that the interplay of these reactions cannot result in an excess of H₂ a priori. The efficiency of H₂ conversion increases with temperature and the use of catalysts.

These processes are all at commercial TRL and applied as mature technologies to produce H₂ from fossil fuels at an industrial scale, with reforming and pyrolysis being the most used in fulfilling the actual H₂ demand. The application of these processes in the reservoir to generate H₂ from the remaining hydrocarbons require the direct injection of heat or air/oxygen to burn the hydrocarbons in situ. Assisted by the EOR background, steam/overheated water injection into a reservoir (up to 350 °C) or the in situ combustion (ISC) of hydrocarbons (up to 700 °C for oil combustion) are claimed to be technical solutions. However, the question remains whether they can convert the remaining hydrocarbons to H₂ economically. The air requirement (AR) is one of the main parameters driving the efficiency of ISC and is defined as the injected air (or O₂-rich gas) amount per oil volume recovered (or burned in the reservoir) as a result of the operation. This also determines the compression capacity, therefore influencing the overall project economics significantly.

Although reservoir temperatures during steam flooding (SF) thermal EOR are generally lower than in ISC, it is also argued that steam injection can generate H₂ by activating some of the above-given reactions like steam reforming and WGS. One of the economic and technical determinants of SF is the cumulative steam/oil ratio (SOR), which corresponds to the steam required to recover a unit volume of oil. This is generally expressed as water volume converted to steam divided by the incremental oil recovery, both given with the same unit. This falls within the range of 0.1 to 8; the lower the SOR is, the more efficient the SF in a given reservoir is.

To expand on the overview of thermal EOR methods that can be used for H₂ generation, some economic features are provided in

Table 3. List of some economic features of thermal EOR and cost of H₂ using state-of-the-art technologies.

Thermal EOR	Cost [USD/kg] 1	Ref.
SF field application (USD/IR, Sm3 oil)	104	[53]
Steam generation (LC of energy 2, USD/t)	20–30	[54]
Steam generation (OPEX, USD/IR Sm3 oil)	4–19	[55]
SF, numerical (USD/IR, Sm3 oil)	108	[56]
ISC air injection (USD/Sm3)	15	[57]
ISC field application (USD/IR, Sm3 oil)	78	[53]
ISC, air cost (Sm3/IR Sm3 oil)	96	[58]
H2 Production	Cost [USD/kg] 3
Blue H2 with 50 EUR/t CO2 4	1.6–2.7	[5,59,60]
Blue H2 with 100 EUR/t CO2 4	3.2–7.2	[5,59,60]
Grey H2 with 100 EUR/t CO2 4	1.5–5.1	[5,59,60]
Green H2 (Electrolysis with wind/solar energy)	1.6–12.0	[5,59,60]
Green H2 (Dark fermentation)	2.0–7.5	[22,33]
Green H2 (Biomass conversion)	1.6–8.1	[22,33]

¹ Data provided by the references (without any inflation adjustment). ² Levelized cost including CAPEX and OPEX. ³ Estimated cost of H₂ production by 2030. ⁴ CO₂ capture, transport and storage costs. IR: incremental recovery; SF: steam flooding; ISC: in situ combustion.

based on the literature data. The same table also includes the costs of H₂ generation with conventional technologies, allowing for a better comparison.

5. H₂ Generation in DGRs

Since the amount of gas remaining after the depletion of a natural gas reservoir is quite small, the technology cannot, a priori, open up economic prospects for H₂ yield. However, overheated water/steam injection into the DGRs is claimed to generate H₂ from the SMR in situ.

5.1. Concept

The technology implies the injection of a catalyst precursor into a DGR at the first stage. Then, the temperature in the reaction zone should be raised to a level at which catalyzed SMR and methane cracking occur based on reactions (2), (3), and (5). The endothermic reactions (2) and (5) require heat input (∆H°^298K + 206.2 kJ/mol and ∆H°^298K + 74.9 kJ/mol, respectively) for the transformation of CH₄.

Due to the expected lower H₂ yield, only a few studies have elaborated the technology. The application success depends on the conversion rate of CH₄ to H₂. Therefore metal catalysts are proposed to mostly break C-H bonds and maintain a high and sustainable activity for a long period of time [51]. Two types of monometallic Ni-based catalysts were evaluated for enhancing the process in an experimental study [13]. The first was an in situ prepared catalyst. It can be delivered into the reservoir in the form of a water solution carrying the catalyst precursor, which is then directly converted through chemical reaction into the actual catalyst in the reservoir. The second was the ex situ prepared nickel-based catalyst. This catalyst can be delivered as a suspension into the reservoir together with steam or overheated water. In the study, the theoretical CH₄ conversion rates (on molar basis) were first calculated for a steam/CH₄ ratio of 5 as given in Figure 3. Both the area of interest for this study (dashed line frame) as well as the P-T range that can be realized in a DGR with steam injection (solid line frame) are highlighted in the figure. As shown, the CH₄ conversion efficiency that can be reached via steam formation in a gas reservoir is quite small, being in the range of 20% at its maximum.

A maximum T of 450 °C was achieved in the study with an optimum P, T, and steam/CH₄ ratio (21 at 207 bar and 450 °C). The maximum CH₄ conversion rate was 5.8% mol. The H₂ concentrations measured in the gas streams (max 12.9% mol) were matching, approximately, the CH₄ conversion rates, confirming the stoichiometry of the related SMR equation. Afanasev et al. also reported that the product gas mixture contains light hydrocarbon gases as well. In one experiment, H₂S was found to make up a significant fraction (up to 8.91% mol) of the product gases. It is most likely formed due to the decomposition of sulfur-containing components of the core (both organic and inorganic) and the interaction of the decomposition products with H₂ synthesized during the experiment (not microbial in origin) [13]. It is mentioned that H₂S acts as a catalyst poison and can significantly reduce the activity of the catalyst even in a small amount. In another work published by the same authors [14], if a Ni-based catalyst was used, it was possible to achieve a methane-to-hydrogen conversion rate of 13.6% in the crushed reservoir cores. However, it is also mentioned that the active zone of the core had to be warmed up to high temperatures achievable only in the case of an effective ISC, which is only applicable in the presence of heavier hydrocarbons in situ. In the literature, higher conversion efficiencies of up to 32% using various catalysts and thermodynamic conditions have been reported [15]. However, it is important to note that such efficiencies have been achieved only in batch (not dynamic) experiments.

Abanades et al. proposed direct methane cracking instead of the SMR process, assessing the experimental reaction characterization in a broad temperature range—from 875 °C to 1700 °C—using a quartz–graphite tube as a porous medium [61]. The process is mentioned to be significantly less endothermic than SMR but with a lower H₂ energy output (484 kJ). However, it presented the fundamental advantage of not producing CO₂ as a byproduct. Carbon atoms are easily stored and could be used for producing special materials, e.g., carbon fibers. Besides that, they contain a lot of energy (390 kJ/mol) that could be exploited in the future. The experiments with and without catalysts were realized in that study. At temperatures above 1350 °C and a pressure of 1 bar, the decomposition of methane is practically complete, and the main component of the produced gas is H₂, with almost no traces of hydrocarbon gases. However, in addition to the high required temperatures, which cannot be generated via SF, the deposition of the solid carbon phase (coke) in porous media is seen to be the main drawback of the process.

5.2. Reservoir Engineering Aspects

5.2.1. Mode of Application

Steam, produced at the surface with a given quality (60–80%), is injected into a reservoir at a predefined rate. Due to changing thermodynamics during the trip in the wellbore and in the reservoir, the steam becomes drier with a potential transition to a superheated state. Heat losses for heating the reservoir lead to steam condensation. The surface facilities and installations required for steam generation and injection are not different from those used in EOR operations, with a steam generator and a pump being the minimum requirements.

In the related literature, it is mentioned that the H₂ separation from other gases is performed by downhole membrane separators. Such membranes made of precious metals (e.g., palladium–vanadium) are expensive because they have only a low filter capacity [18,19]. On the other hand, they are negatively influenced by the presence of other gases like H₂S. Another proposed way of producing H₂ is the separation of the gases directly in the reservoir. It is claimed that the generated H₂ moves upward to the crest of the reservoir whereas co-generated CO₂ accumulates in the deeper parts of the reservoir due to gravity segregation [15]. This idea is worth discussing. The H₂ and CO₂, and probably other gases, are components of the gas mixture generated as a result of the thermal processes in the reservoir. Such a separation of the components of a gas mixture, if any, is a rather slow process. From the related fundamentals of the behavior of gas mixtures, it is known that when gases of different natures are brought in contact (this is the case when they are produced in the reservoir in a process such as that discussed), they do not arrange themselves according to their density but diffuse spontaneously, mutually, and uniformly through each other and thus persist as a mixture for any length of time [62]. Their concentrations can change based on their molecular energy as thermodynamic conditions change over time, a process that takes much longer than economic considerations.

Against this theoretical consideration, it could be argued that the gases are generated in different locations and times so that they find sufficient time to separate from each other and move before complete mixing. This possibility was investigated numerically using a simplified 3D generic model of a DGR. The model of an inclined reservoir with a 1.2 × 1.2 km lateral extent had a thickness of about 100 m while the vertical distance between the top and bottom of the reservoir was 450 m, with a grid cell number of 21 × 21 × 10. The reservoir initially contained CH₄ in a pressure-depleted state (30 bars). Gas generation was modeled using a horizontal well that penetrated through the entire width of the reservoir and was located at the mid-layer of the reservoir. To mimic non-uniform H₂/CO₂ generation, the perforations along the well provided equal masses of 10 t/day CO₂ and H₂ in alternating sequence to the reservoir. This was implemented by modeling each perforation/well–reservoir cell connection as an individual source. The phase behavior of H₂-CO₂-CH₄ was modeled using the Peng-Robinson equation of state provided by the fluid property characterization tool of the commercial reservoir simulator used. The calculated phase behavior was also compared and calibrated with experimental data from the literature [63,64,65]. The H₂ and CO₂ generation rates of 10 t/day were chosen based on economic plausibility. After 10 years of gas generation, the operation ceased and the behavior of the generated gases was observed.

Some results of the study are given in Figure 4. In Figure 4a, the H₂ concentration at the uppermost block of the model reaches a concentration of 0.6 (mole fraction) whereas CO₂ is around 0.05. The CO₂ concentration in the lowermost block reaches approximately 0.5 whereas H₂ is around 0.08 in the same block; the remaining gas consists of CH₄. Figure 4b,c depict the profile of the gas distribution in the blocks with the highest H₂ and CO₂ concentrations, respectively. In conclusion, even in such overstated cases, the mixing occurs with other gases, making the in situ separation of the gases questionable.

One aspect that could contribute to the gas separation process is the individual permeability of different gas molecules through the reservoir rock. This effect is industrially used to separate gases by means of molecular sieves, e.g., in a pressure swing adsorption. However, the pore diameter in commercial zeolite-based molecular sieves is in the range of 3 Å to 10 Å and is therefore significantly smaller than the pores in the reservoir. A few studies claim the use of membrane filters to separate H₂ from other gases in the borehole; however, such efforts are still far from industrial applicability; their capacities are significantly low and their effectiveness is susceptible to the presence of other gases such as H₂S and CO [66,67]. A systematic investigation addressing this aspect of the separation process in the reservoir is missing in the literature.

Another reservoir engineering issue of H₂ generation in DGRs is the time needed to bring the reservoir to the pressure level sufficient to produce the generated gas. With the generation rates due to the discussed processes, the built-up pressure can take decades or even longer.

5.2.2. Economics

Here, a preliminary economic assessment for the application of SF in a DGR is performed for optimistic conditions assuming 1 m³ of reservoir rock with a porosity of 0.2, irreducible water saturation of 0.2, and a CH₄/H₂ conversion factor of 50%, with a steam/H₂ ratio of 1. To heat the reservoir rock with a heat capacity of 1 kJ/kg K from its original T of 100 °C to 400 °C (a minimum required T from laboratory experiments), approximately 250 kg of steam is required. With a remaining gas amount of approximately 2.5 kg CH₄ in 1 m³ of the reservoir and considering the given conversion factor, the corresponding H₂ yield is 80 moles (0.16 kg) of H₂. The levelized cost of energy (LCOE) of steam is 25 USD₂₀₂₃/t at the lowest, including costs of emissions [54,68]. Thus, the LCOE required for the generation of 1 kg H₂ can be approximated with 40 USD₂₀₂₃. Obviously, this amount is strongly dependent on the fuel prices; it can, however, be argued that an in situ H₂ generation price lower than this price cannot be realized as the best case while also considering the optimistic assumptions of the calculations. It should be noted that H₂ generation by the SMR process on the surface costs less than 5 USD₂₀₂₃/kg H₂ including both the capturing of the co-produced CO₂ as well as the subsequent transport and storage into the thematized DGR; this is also referred to as blue H₂ generation (Table 3). The maximum blue H₂ production costs are estimated at 11 USD/kg H₂, assuming the most unfavorable scenarios for natural gas prices [5]. The costs of drilling new wells, capturing CO₂ from the produced gas, and the separation and transportation of H₂ and other OPEX costs are not included in the calculations as it is no longer needed to conclude the non-feasibility from both the technical and economic points of view. Risk costs are also excluded, some of which are discussed below.

5.2.3. Associated Risks

The thermal effects of hot water/steam injection should be considered in the designing of new wells and recompletion of existing wells. Thermal cycling can result in low-cycle fatigue failure in the casing string, either in the pipe body or at a connection. Therefore, it is important to consider the strain capacity of the casing and connections used in related wells [69,70]. The effect of the thermally induced stresses can also be relevant on the geomechanical stability of the reservoir. This is of particular importance considering the confinement issue of the generated gases, e.g., CO₂ and H₂ [71]. On the other hand, if H₂ is stored or produced at the top of the reservoir, as claimed, it becomes more difficult to ensure containment requirements. This is due to the considerably higher mobility of H₂, which has a lower viscosity (approximately two times lower than CO₂ at the relevant pressure and temperature) and higher diffusivity.

There are some other discussions on the challenges of in situ H₂ production in DGRs. One of these is related to the extreme chemical inertness of the CH₄ molecule (due to strong C-H bonds and the lack of polarity), which impedes the CH₄ decomposition process. As a result, it is affirmed that carrying out reaction (4) at a reasonable rate would require an energy input, either in the form of much higher temperatures (>1000 °C) or as electrical discharge (plasma). The use of catalysts as in the above-given example may reduce the maximum temperature of the process but at the cost of added complexity (e.g., difficulty of separating the carbon product from the catalyst) [72].

The key question in the engineering process is “how and why should the in situ H₂ generation in a DGR be competitive to SMR at surface plants including GCS?”.

6. H₂ Generation in DORs

6.1. Concept

Due to its claimed higher potential, the generation of H₂ is addressed more from DORs than DGRs in the literature [16,73]. The in situ reforming of crude oils (heavier hydrocarbons according to reactions 2–6) to generate H₂ necessitates higher temperatures (>700 °C) that can only be achieved with the ISC process in the reservoir; SF cannot deliver such high temperatures. There have been some attempts to enable heavy-oil decomposition at lower temperatures using oil-based nanoparticles in steam application [74].

In ISC, a “fire” is generated in the reservoir and sustained by injecting O₂-rich gas/air continuously. To start the fire in the reservoir, an ignition process is required. The ISC can be applied in forward and reverse modes whereas only forward ISC (or high-pressure air injection, HPAI) is applied commercially. In forward application, the fire front advances in the reservoir and is pushing a mixture of combustion gases, steam, hot water, and mobilized oil towards the production well(s). The forward ISC is further categorized into “dry” and “wet” ISC according to the injection fluid being used to sustain the process: air for dry ISC and air and water for wet ISC. In the reverse process, also called countercurrent ISC, the front moves in a direction opposite to that of air flow [52,75,76].

A typical temperature profile for dry ISC with frontal advancement is shown in Figure 5, depicting the main saturation zones generated by the process. Figure 5 illustrates the advancement of the convective wave whereby points A and B mark the positions of the ISC front and the so-called convection point, respectively. Most of the heat generated during the ISC process is stored/compressed between these two points, also referred to as the burned zone. The incentive to switch to wet ISC is related to the preference to reduce this amount of heat and transfer as much heat as possible to the zone ahead of the ISC front where the oil is.

Basically, ISC involves the oxidation of the crude oil (reaction 4); however, it includes a combination of reactions that take place sequentially and/or simultaneously. These reactions are generally lumped in two groups: low-temperature oxidation (LTO) and high-temperature oxidation (HTO) reactions. Based on ISC experiments, it has been concluded that most of the O₂ is consumed in LTO (around 250–300 °C), forming a coke-like residue. In the HTO mode, which generally occurs above 350 °C, the products of combustion are primarily carbon oxides and water. The fuel for these reactions is the immobile hydrocarbon fraction normally designated as coke formed in LTO. The process requires control over the propagation of the reaction (or combustion/heat) front, which is provided by the proper adjustment of the O₂ supply. This is not straightforward as the reaction front prefers the direction of higher oil saturation (fuel) whereas the injected O₂ tends to flow towards the zones of higher gas permeability/saturation.

There is limited laboratory work that mimics the process of using ISC for H₂ production, and existing studies are far from reflecting real-world reservoir conditions. To date, no field-scale experiment has been published, except for some studies reporting H₂ production during ISC field applications. However, at least one patent exists on the topic [66].

Hajdo et al.’s paper reports the production of H₂ in concentrations as high as 20% (molar basis) in the gas streams of production wells in a Marguerite Lake in situ field application in eastern Alberta [77]. They concluded that the H₂ is generated by the coke gasification reaction (4) followed by the WGS reaction. In their successive papers, Greaves and Xi report experiments of in situ EOR processes also generating H₂ [73,78]. They describe a so-called downhole gasification unit (DGU) to investigate the catalytic steam reforming of naphtha fraction cuts from Statfjord crude oil under high pressures and temperatures. High H₂ conversion rates ranging from 35% to 69% (molar basis), considering the outlet gas composition and remaining gases being CO₂ and CO, are reported due to high P and T and favorable catalyst types, oil compositions, and steam/hydrocarbon ratios. Unfortunately, there is no information available on the material balance of the reactions. Kapadia et al. studied, numerically, the in situ generation of H₂ by the gasification of bitumen from the heavy-oil and bitumen reservoirs in Alberta, Canada [16]. The required reactions were simulated with a reservoir simulator on a generic model by mimicking an ISC process and the authors estimated the H₂ amounts as a result of bitumen burning. Up to 1000 Sm³ H₂ per m³ of bitumen, which burned (approximately 0.1 kg H₂/kg oil) as function of P and T, was reported. In an earlier ISC study on the combustion kinetics of Athabasca bitumen from 1D combustion tube experiments, no H₂ was recorded in the outflow gas composition [79]. In a related patent, the numerical estimation yield of 1 kg H₂ per 64 Sm³ O₂ seemed to be overestimated compared to the yields obtained in the laboratory works [66]. Abu et al. reported three dry ISC tests in a combustion tube carried out at 34.5 bars, with a pre-heating and ignition temperature of 95 °C and 350 °C, respectively, to investigate the performance of regenerated and reused catalysts on the upgrading potential of Athabasca bitumen [80]. Temperature profiles with peaks up to 800 °C were recorded. They emphasized that H₂ generation is due to WGS catalytic and thermal cracking reaching 10% of the total gas production at the outlet. An H₂ generation rate of approximately 1.5 g H₂ per kg of OOIP could be derived from the data provided. A list of the H₂ yields from various experimental ISC works is provided in Table 4. In the table, the proposed H₂ generation reactions, as well as the reported AR values, are also provided. Unfortunately, very few studies reporting H₂ generation have provided AR data; AR values reported in ISC laboratory work and field evaluations converge at 500 ± 100 Sm³/m³ or even higher [81].

Although it is postulated that H₂ occurs in a reservoir as a result of ISC, and this has been observed in laboratory works as stated above, most of the laboratory and numerical studies, as well as the field applications, do not refer to the H₂ production in the effluent gas composition, except, if rarely, as a trace [79,82,83,84,85,86,87,88]. In the early attempts, the lack of analytics may have contributed to the low success rate. If any H₂ was recorded, it was mostly considered an indication of an inefficient process in situ such as low-temperature oxidation or channeling. Additionally, the low or non-existent H₂ observation in an outflow can be attributed to the incomplete and incorrect interpretation of the reactions occurring in the reservoir. It can be concluded that reactions (2) through (6) involve an interplay of various components, with H₂ being both co-generated and consumed during combustion. The rate of H₂ generation during ISC depends on many factors, the most important of which are the thermodynamic conditions, the oxygen content, and the type and quantity of catalyst and the oil.

Table 4. List of the ISC and related experiments reporting H₂ generation: application modes, reported mechanisms, H₂ yields, and applied air consumption (AR) data.

Application Mode	Reported Mechanisms	H2 1	AR [Sm3/m3]	Ref.
Core tests with reservoir bitumen, T = 280–350 °C	Thermal cracking, WGS, coke gasification	1	n.a	[77]
Core sandpack, T = ca. 500 °C	Partial combustion, WGS	2.1	168	[81]
Pilot scale DHG unit, with Naphta cut, T = 750 °C	Catalytic steam reforming	50	n.a.	[78]
Conical combustion cell with Athabaca bitumen	Thermal cracking, LTO/HTO	<5 2	400	[89]
Combustion (sandpack) tube, T = 800 °C	Catalytic combustion, WGS, SG	1.5	220	[80]
Core flood tests, hot water-steam, T = 200 °C	Decarboxilation + the reaction with rock	11	n.a.	[90]
Experimental, 3-D, SADG followed by ISC	Thermal cracking of bitumen	1.2	n.a.	[91]
Conical sandpack with T = 400–500 °C	Partial combustion, WGS	6	n.a.	[57]

¹ Production g/kg oil (of OOIP); ² mol% in total produced gas composition, mass not provided; AR: Air requirement; LTO/HTO: Low-high temperature oxidation; WGS: Water-gas shift; SAGD: Steam assisted gravity drainage.

Table 4. List of the ISC and related experiments reporting H₂ generation: application modes, reported mechanisms, H₂ yields, and applied air consumption (AR) data.

Application Mode	Reported Mechanisms	H2 1	AR [Sm3/m3]	Ref.
Core tests with reservoir bitumen, T = 280–350 °C	Thermal cracking, WGS, coke gasification	1	n.a	[77]
Core sandpack, T = ca. 500 °C	Partial combustion, WGS	2.1	168	[81]
Pilot scale DHG unit, with Naphta cut, T = 750 °C	Catalytic steam reforming	50	n.a.	[78]
Conical combustion cell with Athabaca bitumen	Thermal cracking, LTO/HTO	<5 2	400	[89]
Combustion (sandpack) tube, T = 800 °C	Catalytic combustion, WGS, SG	1.5	220	[80]
Core flood tests, hot water-steam, T = 200 °C	Decarboxilation + the reaction with rock	11	n.a.	[90]
Experimental, 3-D, SADG followed by ISC	Thermal cracking of bitumen	1.2	n.a.	[91]
Conical sandpack with T = 400–500 °C	Partial combustion, WGS	6	n.a.	[57]

¹ Production g/kg oil (of OOIP); ² mol% in total produced gas composition, mass not provided; AR: Air requirement; LTO/HTO: Low-high temperature oxidation; WGS: Water-gas shift; SAGD: Steam assisted gravity drainage.

6.2. Reservoir Engineering Aspects

6.2.1. Mode of Application

A typical ISC is considered. The process has some advantages over SF including higher thermal efficiency, relatively small heat loss to the overburden, and no heat losses in the wellbore, and it can be applied in deeper and high-pressure reservoirs. Decisive factors for the success of the process include the highest possible homogeneous permeability. As previously stated, heterogeneous permeability distribution would lead to unfavorable, uncontrolled fire progression; therefore, homogeneous formations are advantageous. To enable the injection of a catalyst, an alternating injection of water/steam and gas (i.e., air) must occur. The water stands as the carrier of the catalyst, which is added to the injected water in powder form. The process is thus similar to “wet combustion” since water and gas are injected simultaneously. The surface installation at the production side requires separators for separating both the oil, water, and gaseous phase as well as the storage tanks and gas and water treatment facilities. The gas produced has to be further processed in order to capture H₂ from other gases like CO₂, CO, H₂S, etc. Separated H₂ and CO₂ should be conditioned for transport to the final destination. Figure 6 shows a simplified schematic of the surface installations in the case of ISC application to generate H₂ in addition to oil. Due to related experience, ISC application for both EOR and H₂ generation is technically feasible, with a TRL of 5–6 considering that a similar (prototype) system has not yet been demonstrated in the relevant environment. The main challenge is to develop an efficient process for both incremental oil and H₂ generation.

6.2.2. Economics

With some assumptions using values from related laboratory data (Table 3 and Table 4) and with the application of ISC literature, the following economic evaluation can be made.

In Table 4, a maximum H₂ generation rate of 50 g H₂/kg oil is provided as the result of ISC experiments, a value significantly higher than that in any other study. Assuming this yield, a numerical assessment of the total H₂ generation can be made, similar to the above-given cases, for 1 m³. With a porosity of 0.2 and an oil saturation of 0.7, an oil volume of 0.14 m³ is stored in the pore volume. Considering a density of 960 kg/m³ for the heavy oil, about 135 kg oil yields 7 kg of H₂ with the assumption of perfect oxidation in situ. Various references discussing the economic features of ISC provide an approximate value of 150–200 USD/m³ oil, being slightly lower than SF costs [53]. This results in a cost of H₂ production in the range of 3–4 USD/kg. Two crucial aspects of this simplified assessment should be emphasized. The first is that all assumed numbers including the H₂ yield are highly optimistic. The second aspect is the cost of processing H₂ after production, which is not included in this simplistic calculation. Produced H₂ should be separated, captured from the other gas components, and conditioned for transport at the surface, thus significantly increasing both the CAPEX as well as the OPEX. Taking these cost components into account, H₂ production from DOR using ISC technology can be estimated at no less than USD 5/kg. The economic value of the process can be higher if both the incremental oil as well as the potential use of the reservoir for CO₂ storage are considered as opportunities. It is often mentioned that ISC is the most powerful EOR method for heavy oil. Therefore, additional oil recovery can bring the H₂ generation cost to a competitive level. In this case, however, the release of additional carbon emissions and their elimination should be parts of the economic sensitivities. An optimal solution would be a combination of technologies using ISC to produce both H₂ and additional oil, separating the H₂ from other fluids in the wellbore, capturing the CO₂ produced in this process at the surface, and injecting this amount of CO₂ into the same or a neighboring reservoir. Another alternative is to ensure the successive use of reservoir heat for geothermal energy, a case that has been highlighted in recent publications [92]. These scenarios are highly unrealistic given the state of the art.

The above-performed assessment does not include the costs of the risks, particularly considering that the ISC is a highly risky EOR operation compared to other EOR methods.

6.2.3. Associated Risks

The risks are primarily due to the geomechanical issues caused by thermal effects as the reservoir temperature is increased up to 1000 °C [93]. Under elevated T and P, large changes in porosity, permeability, and compressibility occur, which may potentially create leakage pathways to the surface by inducing existing fractures and/or initiating new fractures both in the reservoir as well as in the cap rock. Since only air is injected in ISC, the injection facilities are of less importance than in SF in terms of thermal effects.

If the reservoir rock contains carbonates, the thermal decomposition of carbonates is typical, creating one gaseous and one solid reaction product (CaCO₃↔CaO + CO₂). The reaction is reversible and the equilibrium shifts to the right when the temperature increases due to a thermal operation, e.g., ISC, or when the partial pressure of CO₂ in the gas decreases [94]. These phenomena are crucially risky with respect to maintaining the integrity of the reservoir and cap rock for storing CO₂ in situ.

H₂S is the most reported gas component of ISC operations [77,80,82,87,89,95]. The sulfur content of crude oil and/or of the reservoir is converted to H₂S, which is an HSE issue for field management.

In addition to the risks discussed above, the ISC operation itself is a complex process to operate; the issues in connection with ignition and the up-scaling of O₂ needs encapsulate some of the questions to be clarified specific to the fields under consideration. Over the course of the 60 years until 2013, a total of more than 270 ISC field pilots were conducted, with over 200 of them having taken place in the USA. However, it is likely that only 10–20% of these pilots have reached the commercial phase, of which very few have achieved sustainable success [69,96]. The declining interest in ISC in recent years is probably a consequence of the large number of early failures in field trials. However, most of these failures have been retrospectively attributed to inappropriate reservoir selection [52,75]. In order to minimize the risks, laboratory and numerical studies should be carried out to quantify the above-mentioned critical processes and operating parameters as is common practice in reservoir engineering [16,79,80,97,98].

7. Conclusions

The results of the assessment of the feasibility of H₂ production from DHRs are summarized as shown in

Table 5. Results of the reservoir engineering assessment of the feasibility of H₂ production from DHRs.

Method.	TRL	Cost [USD/kg]	Main Challenges	Perspective
Bio-H2 generation (DORs)	7-8	>25	Low H2 yield; H2S co-generation is the main risk.	Very low to no feasibility
Catalytic methane conversion (DGRs)	5-6	>40	Low yield; high risks at subsurface.	Very low to no feasibility
H2 through ISC (DORs)	5-6	>5	Some higher but inconsistent reported yields; high risks at surface and subsurface. ISC is a challenging technology.	Low feasibility; efficient catalysts and powerful downhole H2 separators can change the situation.

. From a reservoir engineering perspective, the following conclusions can be drawn.

The TRL of bio-H₂ generation in DORs is estimated to be the highest due to the similar applications of MEOR. However, the extremely low H₂ yields limit the economic perspectives of the proposed technology. The same conclusion is valid for the use of DGRs for H₂ generation; in addition, the TRL is lower.

The information provided in the literature on H₂ generation in DORs using ISC is inconsistent. Theoretical approaches indicate the potential for H₂ generation as a result of thermal processes in DORs. However, a significant number of reports and studies in the field, as well as in lab works, disregard the generation of H₂ totally. A small number of studies, including some patents, report the production of H₂ as a result of the potential application of ISC-EOR technology for both light and heavy oils but without economically significant H₂ yields. This may be partly due to the complexity of the experiments and their evaluation and partly due to the interplay of reactions that simultaneously produce and consume H₂. ISC applications are known to be challenging to design and operate. Despite ISC’s potential effectiveness (if successful) compared to SF, there are numerous challenges posed by both methods such as H₂S co-production, material issues, and thermal stresses in the reservoir. On the other hand, the argument of in situ gravity segregation and the production of H₂ leaving heavier CO₂ in the reservoir for geological storage is not readily tenable as a vertical composition gradient is difficult or impossible to achieve within a realistic timeframe.

We conclude that the future of “H₂ production in DOR with ISC” mainly depends on the development of two key technologies. The first is the implementation of suitable catalysts that would increase the efficiency of the process in terms of the H₂ yield. The second is the development of powerful and efficient downhole H₂ filters (membrane separators) that would help keep the other co-generated gases in the reservoir increase carbon neutrality. Industrial efforts continue with ongoing R&D works to close the gap for potential applications [67,99].