Due to accelerated industrialization and increased energy consumption, substantial amounts of carbon dioxide have been released into the atmosphere, resulting in a series of changes in the Earth's climate and weather systems. While countries around the world are actively engaged in carbon dioxide storage projects, these efforts are still insufficient to mitigate global temperature changes (Su et al. 2022; Li et al. 2023; Kumar and Eswari 2023). Since the late 1990s, humanity has started to acknowledge the environmental risks associated with fossil fuels and has shown a growing interest in green energy sources such as solar and wind power. However, renewable energy sources including wind and solar cannot reliably serve as grid-scale power sources due to their intermittent nature unless excess energy can be stored and supplied later during periods of shortage (Jarvis 2015; Sun et al. 2023a). Compressed Air Energy Storage (CAES) is considered a promising solution for mitigating short-term fluctuations in renewable energy production. It achieves this by rapidly increasing energy output and enabling efficient part-load operation (Succar and Williams 2008; Fushimi 2021). CAES system generally includes six main components: (1) compressor, generally multi-stage compressor with intermediate cooling device; (2) expander, generally multi-stage turbine expander with interstage reheat equipment; (3) combustion chamber and heat exchanger for fuel combustion and recovery of waste heat.; (4) storage device, underground or above ground cavern or pressure vessel; (5) motor/generator, connected to the compressor and the expander through the clutch; (6) control system and auxiliary equipment, including control system, fuel tank, mechanical drive system, piping and accessories. As shown in Fig. 1. With advantages such as substantial storage capacity, extended storage duration, high system efficiency, long operational lifespan, flexibility, intermittency management, low cost, and scalability, CAES is regarded as one of the most promising large-scale energy storage technologies (Ozarslan 2012; Wan et al. 2023a; Wang et al. 2018).

Working principle and main components of a CAES station

These facilities typically take two primary forms:

aboveground liquefied natural gas (LNG) ball tanks and underground gas storage (UGS) (Liu et al. 2014). UGS encompasses various types, including gas reservoirs, oil reservoirs, salt caverns, and abandoned pits (Cooper et al. 2011). Notably, more than 75% of the world's gas reservoirs are currently of the depleted reservoir type, and 81% of globally stored underground natural gas is found in depleted oil and gas fields (Xie et al. 2009). CAES brings economic benefits by using depleted, hydraulically fractured oil and gas wells to store electrical energy in the form of compressed natural gas. The porous geologic environment of fracked wells, which is used to release hydrocarbons, is also conducive to storing and releasing gas on a daily or seasonal basis. Round-trip storage efficiencies are estimated to range from 40 to 70%, based on natural reservoir temperature, with storage costs estimated at $70–270/MWh, making it comparable to pumped storage (Young et al. 2021).

The United States was the first country to show interest in CAES technology, with the publication of the first literature paper in 1976. However, Europe (EU-27) and China lead the way in research on CAES today, whose publication accounts for nearly half of the published literature on the subject. Notably, China has become the leading investor in CAES development, with several demonstration and commercial CAES plants currently under development and commissioning (Borri et al. 2022).

The types of gas storage include salt cavern, depleted oil and gas reservoir and aquifer. The surrounding rock of salt cavern has good creep property and the high salt content can inhibit some microorganisms, but the suitable sites are few and the gas storage is limited. Aquifers have large gas storage capacity. However, they have long construction period and high cost. The produced gas also needs to be dehydrated.

On August 17, 2023, the international first 300 MW-class advanced CAES system expander jointly developed by the Institute of Engineering Thermal Physics (IETP) of the Chinese Academy of Sciences and China National Energy Technology Co., Ltd. completed the integration test and successfully rolled off the production line. Its successful development will promote China's advanced CAES technology to a new level (Hou 2024; Agrawal et al. 2023; Yang et al. 2023a, b).

In recent years, more research has been conducted on the application of gas storage, including ground hydrogen storage for zero emissions (Al-Yaseri et al. 2023; Peng 2023; Nguyen 2023; Kalam et al. 2023) and carbon capture utilization and storage strategies for carbon dioxide injection (Gao et al. 2023a, b). The development of depleted oil and gas type reservoirs is of great significance to the change of energy structure and the promotion of the development of energy technology, and also lays a solid foundation for the construction and development of smart grids, energy internet and smart cities (Feng 2023). Urgent verification is needed for energy storage feasibility, for this reason, this paper combines the development history of CAES technology to research on the site selection of depleted gas reservoirs (DGR), reservoir dynamic sealing evolution mechanism, and high flow rate CAES injection and extraction technology, to support the development of the depleted gas storage type reservoirs.

The overall goal of CAES is to store energy during periods of low power demand and then use it during periods of high demand. Conventional CAES satisfy the following concepts, excess electricity is utilized to compress the surrounding air, capturing and storing heat in a thermal energy storage system, which is applicable for adiabatic CAES. The compressed air is stored in a vessel, later released and preheated in a heat exchanger, and directed to a turbine generator to produce expensive electricity. Finally, the electricity is fed back into the electricity grid (Duhan 2018; Sun et al. 2023b; Zong et al. 2023).

2.1 Current status of CAES technology

2.1.1 Principles of operation

There are many types of CAES technologies, which can be classified into three categories according to whether they require preheated air in the combustion chamber, the size of the storage, and whether they utilize heat of compression as shown in Table 1 (Xu and Song 2021).

Figures 2a, b show the schematic diagrams of the supplementary fired CAES and non-supplementary fired CAES. The supplementary fired CAES system is based on the working principle of gas turbine, the supplementary fired chamber is set up at the entrance of the turbine, and the fuel is utilized to heat the air to increase the amount of work done by the turbine, which is a reliable and stable system. However, the fuel combustion will emit pollutants and cause environmental pollution, which does not conform to the requirements of the development of green environmental protection. The non-supplementary fired CAES system abandons the traditional compensatory fired chamber and utilizes a heat storage device to collect the compression heat generated during the air compression process, which is used to heat the inlet air of the first-stage turbine expander when releasing energy, thus realizing a zero-pollution working process (Xu and Song 2021; Li et al. 2022).

The difference between supplementary fire and non-supplementary fired systems. a Supplementary fired system b Non-supplementary fired system

Isothermal CAES systems use certain measures (such as pistons, showers, bottom injection, etc.), through the specific heat capacity of the liquid (water or oil) to provide an approximate constant temperature environment, increase the air–liquid contact area and contact time, In this way, the air in the process of compression and expansion is infinitely close to the isothermal process and the heat loss will be reduced to a minimum, improving the efficiency of the system (Dolatabadi et al. 2013; Hu et al. 2023).

Due to the high technical requirements and costs associated with the realization of single-stage adiabatic CAES concepts, great interest has been shown in Isothermal or Quasi-Isothermal CAES concepts in recent years. As shown in Fig. 3, the focus here is on dividing compression and decompression into several stages so that each stage is associated with only a slight temperature increase (Donadei and Schneider 2016).

Sketch of the principle behind the isothermal concept

Statistics on some isothermal CAES systems at home and abroad, as shown in Table 2. The isothermal CAES systems mainly use liquid piston technology, and the average circulation efficiency is about 70%.

In addition to the basic types mentioned above, researchers have proposed many CAES derivatives based on the fundamentals of CAES in the innovation stages. Among them, the more representative derivative schemes are:

1. (1)

   Liquid air energy storage (LAES)

As shown in Fig. 4, according to the liquefaction phase change properties of air, compressed air is liquefied and stored in low-temperature storage tanks. As the density of liquid air is more than 10 times that of CAES, the container volume required for air liquefaction storage will be greatly reduced, reducing the impact of geographical conditions, but its conversion efficiency needs to be improved (Morgan et al. 2015).

Liquid CAES system schematic

1. (2)

   Super critical compressed air energy storage (SC-CAES)

As shown in Fig. 5, its components and the existing CAES system and liquefied air energy storage system is more similar. It can be used as a heat and cold storage device for air compression. At the same time, which not only has much higher energy density than that of CAES, but also greatly improves the efficiency of LAES (Liu 2012; He et al. 2018).

SC-CAES system schematic

1. (3)

   Small scale CAES (SS-CAES)

Small scale CAES system has less requirements for the geographic location, and it can be used in the form of tank storage of compressed air storage. In order to maintain a constant temperature and high-pressure safety of tank, it can be buried in the ground, and the efficiency of this system can be up to about 50% (Xu et al. 2021) (Table 3).

2.1.2 Development status

The development of CAES technology is inseparable from the change of energy structure, which can be roughly divided into three stages: rapid development, slow development, and then rapid development. Since 1949, the German engineer Stal Laval put forward the concept of energy storage using compressed air in underground caverns. Each country has carried out a lot of research and practice. The world's two earliest CAES systems, were established in Germany in 1978 with a power of 290 MW Huntorf CAES system, as well as the United States in 1991, the power of 110 MW McIntosh CAES system (Guo 2013; Budt et al. 2016). The parameters of these two systems are shown in Tables 4 and 5.

Combining the actual circumstances of oilfield enterprises, utilizing underground porous media space to rebuild energy storage can reduce the cost of electric power consumption in oil and gas fields and improve production efficiency. This is one of the best paths to realize the synergistic development of "energy storage" and "underground resource utilization". Domestic oilfield enterprises such as Shengli Oilfield, Daqing Oilfield, Qinghai Oilfield, and Jilin Oilfield have already deployed plans to convert depleted gas reservoirs into energy storage and have conducted preliminary exploration. In June 2022, Shengli Oilfield completed the project in collaboration with Tsinghua University and other units. In April 2023, Qinghai Oilfield conducted bidding for the horizontal well fracturing and construction of CAES project. The other projects are shown in Table 6.

After China completed the 0.5 MW Wuhu non-supplementary fired demonstration project in 2014, the 10 MW CAES validation platform in Bijie, Guizhou and the 10 MW CAES peaking power plant in Feicheng (Phase I) went into operation in 2021, and the 100 MW CAES project in Zhangbei entered the power-carrying commissioning stage in 2022 with the technical support of the IETP (Zhao et al. 2023).

The development history of CAES projects is shown in Fig. 6 The earliest program was Stal Laval in Germany in 1949, followed by the rapid growth of the CAES program in China in recent years. With the support of Tsinghua University, the 100kW composite CAES industrial demonstration project in Xining, Qinghai was put into operation in 2016, the 60 MW salt cavern CAES in Jintan, Jiangsu Province has been connected to the grid for power generation in May 2022 (Guo et al. 2019). In addition, the projects of Hubei Yingcheng 300 MW, Gansu Jiuquan 300 MW, and Shandong Tai'an 350 MW, under China Energy Digital Technology Group Co.,Led., have already started construction.

CAES project development history

2.2 Key equipment for CAES

As shown in Fig. 7, CAES system contains compression, gas storage, heat/cold storage, heat/cold return, expansion power generation and other sub-systems. The key equipment mainly includes compressors, heat exchangers and expanders and the technology of the relevant equipment is relatively mature. Through the project demonstration and construction, it has a certain industrial chain basis.

Depleted gas reservoir CAES working principle and its main equipment

2.2.1 Compressor

Compressor, mainly divided into turbine, piston and screw type, is a kind of compressed gas to increase gas pressure or transport gas machine. It can be used for CAES system compressor and has the characteristics of large flow rate and high pressure. Currently, there are more manufacturers of compressors on the market and the technology is more mature. The main manufacturers include Atlas Copco, Comp Air, Sull Air and Siemens, etc. Atlas Copco has the most mature air compressor manufacturing technology and the highest global market share. Comp Air is a leading supplier of world-class rotary screw, reciprocating, centrifugal, and portable compressors. Sull Air Compressors originated in Michigan City, Indiana, USA, and has specialized in the development and manufacturing of screw air compressors for more than 50 years, with the following product types: stationary oil-flooded, stationary-oil-free, vacuum pump and other product types.

In China, China National Petroleum Corporation(CNPC) Jichai Power Company Limited has a wide range of products with power ranging from 10 to 7500 kW and maximum working pressure of 70 MPa, which can meet the different needs of natural gas booster and gathering and transportation.

The parameters of the two compressor systems are shown in Table 7, where the maximum discharge pressure reaches 52 MPaG.

1. (2)

   Heat exchangers

Heat exchangers are mainly categorized into shell and tube type and plate type. They are heat exchanger equipment that transfer part of the heat from the hot fluid to the cold fluid. Among them, the parameters of heat exchanger system have a greater impact on the energy storage efficiency of the system. If the heat storage temperature and heat return temperature are higher, there will be lower loss of the system and higher energy storage efficiency of the system.

By improving the heat storage temperature and heat transfer efficiency of the heat storage and return system, the overall efficiency of the system can be further improved. ARD's heat exchanger production technology is more mature, and its main products include detachable plate heat exchanger, heat exchanger gasket and heat exchanger plate.

1. (3)

   Expansion machines

Expansion machine according to the structure and form of movement can generally be divided into turbine and piston type. It can be used to compress the gas expansion and decompression and output power to the outside, so that the temperature of the gas is lowered. Piston type is mainly suitable for small flow and high-pressure ratio of small and medium-sized high and medium pressure cryogenic equipment, while the turbine type has a small size, simple structure, high flow, high efficiency and long operating cycle, etc., suitable for large and medium-sized deep cryogenic equipment. The turbine expander is generally used in CAES system.

Now in its 150th year, Atlas Copco's centrifugal turbine compressor solutions utilize either integral gear drive technology or single-shaft drive technology, and are capable of handling pressures up to 20 MPa and volumetric flow rates up to 480,000 m³/h (Quoilin et al. 2012).

Atlas Copco has a wide range of expansions with the product characteristics shown in Table 8, and its maximum inlet temperatures is up to 510 °C.

The requirements for CAES site selection in DGR mainly include four aspects: reservoir geological conditions, geologic safety, historical factors, and economic efficiency factors. The CAES project for DGR conducted by PG&E in California, USA, is taken as an example for analysis. A detailed analysis is carried out on it, and its comprehensive evaluation system for CAES is shown in Fig. 8 (Jia et al. 2015).

Comprehensive evaluation system for CAES

The main purpose of the construction of pressurized gas storage power plants is to regulate the peaks and valleys of electricity and to improve the quality of electricity. Regulating power peaks and valleys is to alleviate the difference between day and night peaks and valleys in the power market, and to regulate the balance of power consumption in the power grid in time and space.

The improvement of power quality is mainly aimed at improving the unstable quality of intermittent power sources such as wind power and photovoltaic power generation, and storing a large amount of wind and light discarded power during the peak hours of grid supply. The construction purpose of the pressurized gas storage power plant is the primary factor in determining the regional siting of the storage reservoir, while the force characteristics of the underground storage reservoir are the key factor in determining whether it can be successfully sited. The selection principle of CAES is shown in Table 9.

3.1 Evaluation of reservoir property factors

3.1.1 Reservoir size and thickness

Reservoir size can be evaluated based on two dimensions: reservoir capacity and extent of distribution. The selection, of reservoir size depends on the storage and operational needs of the system as well as the value it can generate. Reservoirs that are too small do not have sufficient capacity to sustain gas recovery operations to meet project objectives, and reservoirs that are too large require the construction and maintenance of larger gas tops, which increases development and operating costs.

It is promising to utilize energy storage in coal goaf under the current "dual carbon" target (Wang et al. 2023b). In the context of underground coal seam gasification reactions, Greg Perkins proposed a 0-dimensional cavity growth sub-model based on the concept of surface reactions, which provide more accurate cavity growth rates with reasonable input parameters (Perkins 2019).

Reservoir thickness is usually expressed as gross, net and average thickness. Gross thickness is measured from the highest point in the reservoir structure to the gas or water contact that normally defines the lower limit of the reservoir, without regard to changes in lithology within that interval. The net thickness is derived from log interpretations and excludes those lithologies within the reservoir that are of poor quality. A higher average reservoir thickness may imply a relatively compact reservoir spread compared to gross thickness, and conversely, a lower average thickness may imply a wider and more extensive reservoir spread.

3.1.2 Reservoir depth and pressure

When selecting the site, the depth of the reservoir should be neither too small (small circulating pressure causes low energy efficiency and requires more storage space) nor too large (the pressure is limited by the safety of the system, the economy and the performance of the equipment) (Dong and Li 2021). The burial depth of the reservoir primarily determines the range of pressure variations in the CAES system during buffer gas injection and cycling.

Different pressure ranges have a significant impact on the overall design of the reservoir, the surface compressor and the expander. When the effect of depth on compressor and expander design is not considered, the greater the depth of the target reservoir, the greater the storage efficiency of the overall storage system. And the geothermal energy is more likely to be recharged from the surrounding strata. However, with the increasing depth, the pressure buildup from injecting buffer gas is greater, which may cause mechanical damage to the storage cap layer.

Tables 10 and 11 give the range of reservoir depths for selected projects and the range of reservoir depths considered by the researchers in their evaluation of CAES siting.

3.1.3 Reservoir porosity

The pore structure of the reservoir directly affects the physical properties of the reservoir and has an important influence on the reservoir storage and seepage capacity (Wang et al. 2021). The porosity of the reservoir reflects the size of the pore volume of the rock, which can be interpreted from bare-eye logging curves or indirectly obtained from core analysis. Stottlemyre and Allen et al. in 1978 and 1983, respectively, proposed a reservoir porosity of greater than 10% for the siting of CAES in aquifers, and Succar et al. in 2008 proposed a minimum porosity of 13% for the reservoir (Dong and Li 2021; Ngata et al. 2023).

3.1.4 Reservoir permeability

For low-permeability reservoirs, small changes in formation pressure will cause changes in reservoir porosity and permeability, which in turn affects the seepage capacity of the underground reservoir and ultimately affects the amount of gas injected into the underground reservoir. The lower the permeability of the reservoir is, the more drastic the change in permeability with the formation pressure will be (Guo et al. 2021; Zhang et al. 2019a). Permeability needs to be determined by analyzing core samples from the reservoir to determine actual vertical and horizontal permeability. The effect of permeability and reservoir thickness on reservoir performance needs to be assessed through reservoir modeling. Zhang et al. (2022) studied the injection and extraction simulation of low permeability gas reservoirs converted into underground storage reservoirs. Based on the inverse problem theory, the objective function was constructed by using the difference between the measured and calculated values of formation pressure, and the problem of inverse identification of reservoir physical parameters was transformed into an optimization problem (Stottlemyre 1978; Hostetler et al. 1983).

3.1.5 Surplus oil and gas

All DGR contain a certain amount of residual gas, which can be estimated by analyzing the historical production data of the reservoir to estimate the original gas-in-place. Over time, the residual natural gas will gradually mix with the injected air, and the lower flammability limits of the methane-air fraction of a representative deep geologic reservoir are 3.8 mol% and 54.4 mol% at 25 °C and 85.5 atmospheres, respectively (Zhang et al. 2022; King and Apps 2013).

3.1.6 Trapping mechanism

As shown in Fig. 9, for depleted oil and gas fields, the effectiveness of the trap depends on (1) whether the burial depth and area of the trap are favorable for the economic construction of the reservoir, (2) whether the closure height of the trap, and the cap and faults in and around the trap are favorable for the preservation of the injected natural gas, (3) the effect of the trap overflow point and the pressure of the overflow on the escape and transport of the stored gas (Zheng et al. 2020).

Schematic diagram of geological body of a depleted oil/gas field UGS (Zheng et al. 2020)

According to foreign experience, simple formations such as backslopes or fault traps are easier to develop and operate than complex formations. The more complex the reservoir becomes, the more likely it will be to need additional wells. It will be more difficult to operate due to connectivity barriers within the reservoir.

3.1.7 Number of production layers

Some oil and gas fields are exploited with multiple producing horizons at different depths, which may be connected leading to mixing of oil and gas, or may be unconnected. Therefore, it is impossible to calculate the actual size of the reservoir based on the production records of a particular production formation. At the same time, the reserves of a single producing formation do not meet the economic efficiency and production requirements of CAES, resulting in increased construction costs. To predict the geological reserves, firstly, the geological structure, reservoir condition and obtained oil and gas layers of a certain block should be counted, and the reservoirs with clear stratification will provide certainty for the reservoir size and the convenience of development.

3.1.8 Drive mechanism

Gas reservoirs are usually gas-driven and water-driven for gas recovery, and gas-driven reservoirs usually have very high gas recovery rates, above 80%. To study the driving mechanism, it firstly should start from the microscopic pore characteristics of the reservoir, and the experimental testing methods such as physical property test, cast thin section, scanning electron microscope, high pressure mercury pressure, etc., should be used to classify and study the petrological characteristics, physical properties and microscopic pore structure characteristics of the cores taken from this reservoir. On this basis, the oil–water two-phase seepage experiments are carried out through the sandstone model to reflect the pore structure (He et al. 2020; Xiong et al. 2023).

3.1.9 Geological complexity

From the perspective of CAES development and operation, the simpler reservoir geology is better, so as to reduce potential development risks and save costs. Prior to the development of a depleted reservoir, the geology needs to be comprehensively evaluated combined with regional geological interpretations and logging records from exploration and production wells.

3.1.10 Types of reservoir native minerals

The chemical reaction of primary minerals with air in different reservoirs can have an impact on the economics and safety of the reservoir. Geochemical element logging, which provides access to the mineralogical composition of a reservoir, is a method of obtaining geochemical element data to determine the mineralogical composition of a tight reservoir. It is identified primarily by detecting gamma rays produced in the formation by neutron reactions. Oxygen from compressed air entering the reservoir will oxidize with the primary minerals in the reservoir. Newly generated oxides can reduce the permeability and porosity of the reservoir due to increased volume or precipitation, as well as reduce the oxygen in the output gas, affecting the combustion efficiency of the fuel that enters the fired chamber during subsequent power generation. Therefore, when site selection is carried out, the primary mineral type of the reservoir can be analyzed through geochemical elemental logging, and reservoir areas with primary mineral types of iron or calcium with high sulfur content can be avoided as much as possible (Liu 2022).

Ying et al. (2023) established a multilayer model based on the fluid properties of rocks in the Huangcaoxia gas field. In order to understand the removal process of H₂S from sulfur-containing UGS, the evolution law of H₂S in the underground reservoir of Huangcaoxia sulfur depleted gas field was simulated by the numerical simulation method.

3.2 Geological safety factors

The sealing and stability of the geologic structure plays an important role in the safety of the entire energy storage system. When evaluating the site selection for underground CAES, whether the whole site can be used for CAES, and the safety and stability of its energy storage system must be considered (Vandeginste et al. 2023).

3.2.1 Cap stability

For a gas storage reservoir, the capping capacity of the cap is the ability of the reservoir to prevent the escape of natural gas, which controls the vertical distribution, abundance, and working pressure of natural gas in the reservoir (Liu et al. 2021). The CAES process is prone to pressure buildup in the reservoir due to the need to inject a large amount of gas into the reservoir, and the excessive pressure may destabilize the cap layer. The stability evaluation of the cap layer is mainly a study of its geo-mechanical properties. Mechanical effects of the cap layer may cause opening of initial fractures in the cap layer, and destruction of the cap rock or rock mass. Higher injection pressures may make the cap layer incomplete and induce potential leakage channels. The geo-mechanical stability of the cap layer can be investigated by analyzing the stress–strain relationship of the rock through triaxial tests (Bai 2008).

3.2.2 Site stability

Site stability evaluation mainly refers to the evaluation of the impact of geological tectonic movements or natural disasters on the gas storage structures and ground supporting engineering facilities at the candidate site. Earthquakes and active faults can drastically disrupt the confinement conditions of the storage system, thus affecting the stability and safety of the entire storage system. Among them, fault activity has greater impact. The active faults are the main source of risk for destructive earthquakes and the main cause of near-surface tectonic deformation, and their existence implies potential and unpredictable earthquakes, surface deformation and related secondary disasters and hazards (Xu et al. 2012; Wu et al. 2022). Combined with the geological complexity, the stability of the sites is evaluated and people prioritize these sites without the above risks to avoid major safety accidents.

3.2.3 Cap tightness

The macroscopic influencing factors of cap tightness mainly include lithology, thickness, burial depth, distribution continuity, mechanical stability, fault and fracture development and closure. The evaluation of aquifer compressed air storage for the closure of the cap layer can draw on the evaluation method of cap layer confinement in oil and gas engineering and CO₂ geological storage engineering to a certain extent (Diao et al. 2011; Wang et al. 2023a).

3.3 Historical factors

Historical factors include human intervention for the reservoir, such as exploration and production wells, gas production from the target reservoir and surrounding reservoirs, and well plugging and abandonment. Four types of historical data from prior exploration and development efforts are also important for CAES development, including the number and type of wells in the gas reservoir, well age, and abandonment history.

3.3.1 Number and type of wells

The advantages of using depleted reservoirs for energy storage are the availability of detailed geological information and historical production records, lower exploration costs and shorter construction periods. According to statistics, the number of abandoned wells worldwide exceeds 20 million, and in 2023, the number of abandoned wells in China has exceeded 100,000 (Raimi et al. 2021; Fang 2023). If there is greater number of wells, there will be greater potential problems and higher additional development costs. When selecting a site, a variety of factors should be analyzed, including the location, number, type and production history of each well, eliminating possible production hazards. Although no specific screening criteria is established, priority is given to reservoirs with fewer wells drilled.

3.3.2 Well age

Older wells have a higher risk of failure and potential leakage and they need to be analyzed for the economics of converting an underground storage reservoir by learning the quality of cementing of old wells, plugging of abandoned wells, underground information, infrastructure (water, electricity, transportation, etc.), and planning for new wells to be drilled. If more underground information and surface infrastructure are available, there will be lower number of old wells to be rehabilitated and new wells to be drilled, and it will be more economical to build the reservoir (Jia et al. 2016).

3.4 History of abandonment

The abandonment history of each well needs to be evaluated to ensure that potential hazards such as leaks do not occur. Many wells located within a target reservoir have been abandoned at the time of drilling or after a period of production. The following records of abandoned wells need to be evaluated: the geographic location of the abandoned well, the type and quantity of cement or other waste materials, the abandonment process, the time the well was abandoned and mined and the presence of foreign material in the abandoned well.

3.5 Economic benefit factors

As shown in Table 12, the construction cost of the three domestic gas storage tanks is the lowest in Jintan, which is mainly due to the use of existing caverns to reduce the construction cost, and has a large volume so that the cost per unit volume is the lowest.

As shown in Table 13, the main factors affecting the economic efficiency include, geographical location of the site, investment cost, scale of energy storage and economic loss due to wellbore corrosion in four evaluation indicators.

The engineering background of multi-cycle intensive injection and extraction in the lower reservoir makes the dynamic sealing capability and stability of trap caps and columns under alternating loads a critical issue that cannot be ignored (Wen et al. 2021). When the gas breakthrough enters the cap layer, the gas replaces the water in the voids or fissures, and at the same time, the mechanical effect of the cap rock will be changed with the gas breakthrough. If the conditions of microcrack expansion of the cap rock are reached, the length and tensile degree of microcracks will be increased, and the permeability will rise sharply (Peng et al. 2014). The capping layer is depicted in different time scales as shown in Fig. 10.

Schematic diagram of the multi-field coupling process inside the cap layer

4.1 Capillary sealing of capping layer under alternating stresses

4.1.1 Capillary closure mechanism

There are many parameters for evaluating the capillary closure capacity of the cap layer, mainly the replacement pressure, permeability, porosity, surface area ratio and microporous structure of the cap layer. However, according to statistics, it is found that there is an obvious functional relationship between the parameters of cap porosity, permeability, specific surface area, microporous structure and median radius of voids and replacement pressure, which indicates that the role played by these parameters in the evaluation of the capillary closure can be replaced by the replacement pressure (Davies 1991).

At present, the breakthrough pressure is mainly used to quantitatively evaluate the capillary closure, which is the most fundamental and direct evaluation parameter of capillary sealing ability of the cover layer. It comprehensively reflects the influence of lithology, mud content, porosity, permeability, microscopic pore throat distribution on the capillary closure (Wen 2021).

4.1.2 Changing law of permeability of cover layer under alternating loading

Rock permeability is controlled by the pore and fissure structure of the rock itself, and during the deformation process, the pore and fissure of the rock changes, and therefore its permeability also changes. Rock damage itself is an extremely complex problem, and it is difficult to study the seepage-stress coupling in the damage process theoretically, and the main way to study it is to conduct experimental research.

At present, there are more experimental studies on the permeability of mudstone before the peak, and the conclusions are basically the same: in the elastic deformation stage, the mudstone micropores and fissures are compressed, and the permeability decreases; with the further increase of stress, the rock micro-fissures begin to expand, and the permeability begins to increase; the peripheral pressure restricts the lateral deformation of the mudstone, which reduces the porosity, and restricts the fissure expansion and width. With the increase of the peripheral pressure, the permeability decreases (Han et al. 2011).

4.2 Mechanical integrity of the formation under variable stresses

Those factors including the tensile damage of reservoir and cap layer caused by local high pressure under alternating stress, and the risk of cap layer shear and long-term fatigue damage caused by local stress concentration due to complex geological structure, lithological changes, and laminar development, etc. are the focuses of formation mechanical integrity evaluation.

4.2.1 Mechanical closure mechanism of the cap layer

The capping layer must be thick enough to prevent rupture, and it needs to have low permeability and large capillary forces to prevent air migration through the capping layer. According to experience, the injection pressure in excess of the original formation pressure should not exceed 0.16 bar/m depth of burial to avoid cracking of the capping layer (Succar and Williams 2008; Liu et al. 2021).

4.2.2 Damage mechanism of rock deformation under alternating loads

In situ rocks are essentially subjected to monotonic and cyclic or dynamic loads. Correct and detailed knowledge of how the mechanical properties of rocks change under different loading scenarios is necessary for the safe and correct design and construction of civil, mining and geotechnical structures (underground openings, tunnels, rock columns, foundations), as well as for a better understanding of other related operations (Vaneghi et al. 2018; Chen et al. 2023; Zhang et al. 2023).

Liang et al. (2019) conducted a study on acoustic emission-based damage and fractal evolution trend of sandstone under loading and unloading conditions of isotropic layered cycling. The damage variable increased sharply in the cycling phase, and the increment of layered cycling was higher than that of isotropic cycling by 0.07. Sandstone showed greater damage under the action of layered cycling loading and unloading.

Zhang et al. (2021) established the damage ontology model of reservoir and cap layer through triaxial cyclic loading and unloading synchronized permeability test experiment, and studied the strength, permeability changing law and damage law of rocks under cyclic loading. The research results show that: under the action of external force, sandstone is easier to form connecting cracks and be damaged, while mudstone is not easy to produce connecting cracks due to the reduction of permeability by hydration and expansion.

1. (1)

   Risk of tensile damage

High-rate injection and extraction during reservoir operation can exacerbate the effects of reservoir non-homogeneity. Especially during gas injection, the bottomhole pressures may exceed the upper limit of reservoir design pressure. The local pressure may be higher than the minimum horizontal stress, causing tensile damage to the cap layer. In particular, the risk of tensile damage is much higher than the risk of shear damage in gas storage reservoirs modified by shallow buried reservoirs. Therefore, when evaluating the risk of tensile damage to the cap layer, it is important to accurately test the captive ground stress, especially in DGR. The reservoir and cap geos-tresses can be tested by hydraulic fracturing or ground leakage tests and AE Kaiser effect experiments to evaluate the risk of tensile damage (Zheng et al. 2017).

1. (2)

   Risk of shear damage

Evaluating the shear damage of the cap layer is mainly based on indoor rock mechanics experiments. Numerical simulation is carried out by establishing a three-dimensional dynamic geologic force model, and data are obtained by inversion (Teatini et al. 2014). On this basis, according to the shear damage criterion (e.g., Moore-Cullen criterion), with the shear damage safety index and other quantitative indexes, the cover shear damage safety index of the reservoir under any formation pressure during the injection and extraction process is calculated, so as to quantitatively evaluate the risk of the shear damage of the reservoir under the local high-pressure gas injection and the long-term alternating loads (Sun et al. 2017).

1. (3)

   Risk of fatigue damage

The ground stress field in a gas storage reservoir varies cyclically with the injection and extraction cycles. In addition to varying degrees of elastic–plastic deformation, localized stress concentrations may be induced, and such stress concentrations can accumulate in the rock and form fatigue damage. Fatigue damage begins where the stresses are higher and will eventually lead to fatigue damage once microscopic deformation of the tissue begins to accumulate (Ren et al. 2019).

The fatigue damage risk evaluation is to carry out indoor core triaxial loading and unloading alternating stress experiments to study the deformation and damage characteristics of the cap rock under simulated gas storage reservoir injection and extraction working conditions. It quantitatively evaluates the fatigue damage risk of the cap layer under the long-term alternating loads of the gas storage reservoir by using the accumulated plastic strain (Tenthorey et al. 2013). Ma et al. (2018) established a method to carry out triaxial loading and unloading alternating stress experiments by using constant circumferential pressure and variable axial pressure, and recommended the use of 0.1 Hz as the loading frequency of alternating stress for core experiments.

1. (4)

   Fault shear-slip instability

Fault shear slip instability is similar to the principle of cap shear damage, but the fault is a geologically fractured zone, which is the largest mechanically weak surface, and cohesion is generally ignored. Geomechanical studies show that in the process of ground stress disturbance caused by gas storage reservoir injection and extraction, when the shear stress acting on the fault surface is greater than the product of the friction coefficient and the effective positive stress, the fault slips and loses its sealing ability. Slip and destabilization of far-field faults can cause substantial deformation of the formation, which in turn affects the integrity of the wellbore (Zheng et al. 2017).

The primary focus of high-flow pressurized gas storage is on pipe column safety and the study of injection and extraction schemes. Currently, international research on utilizing depleted oil and gas reservoirs for gas storage is still in the exploratory and theoretical analysis stage. The Pacific Gas and Electric Company (PG&E) in California, USA, has developed a mature technology research program for a 300MW-10h scale CAES plant located in the King Island depleted gas field in San Joaquin County (U. S. Department of Energy 2013; Wu 2019).

5.1 Safety evaluation of high-frequency injection pipe column

During high-frequency injection and extraction, the injection and extraction pipe column is affected by corrosion and stress, and is prone to fatigue damage and fracture failure, which has a greater impact on the safety and stability of the gas storage reservoir (Wan et al. 2023b). In the design of the injection process and completion pipe column, it mainly relies on optimization and design software. It can comprehensively consider the factors such as load change, temperature and pressure alternating influence and corrosion, which have a great influence on the safety of the tubular column in the design process. The current research on column corrosion is more mature, and relevant corrosion prediction models have also been established. However, due to the dominant CO₂ and sulfide corrosion in the field, the study of oxygen corrosion is more focused on the study of the process of oil drive with the injection of air. There is less study on the separate injection of air, the corrosive influence of different flow rates, temperatures and pressures on the entire injection column.

Oliveira et al. (2021) introduced an additional module in the MATLAB reservoir simulation toolbox and described a new quadratic method, defined as the derivative of a stratigraph-modified Lorentz diagram, which is based on the flow unit velocity at these depths. The depth range in the reservoir is divided into barrier, strong baffle, weak baffle and normal unit, and the ability of the analysis module in observing well geology is verified by case study.

Yao (2021) evaluated the risk of gas storage pipe columns by fuzzy comprehensive evaluation method. For the establishment of the evaluation index system and the calculation of the index weights in the evaluation process, they chose the fishbone diagram and the hierarchical analysis method respectively, and set up the fishbone diagram-fuzzy hierarchical comprehensive evaluation model to judge the risk level of the pipeline columns, deriving the relative importance of the influencing factors for the damage of pipeline columns.

5.2 Research and optimization of efficient injection and production scheme

At present, most of the domestic and international studies on CAES projects in DGR are at the stage of theory and field test, and there are fewer studies on efficient injection and extraction programs. In the CAES operation process, the safety of injection and extraction column of reservoir can be evaluated by using methods including software and indoor experiments. DGR storage can meet the flow rate and pressure required for system operation, but proper system management and operation must be performed. During the initial gas top formation phase, additional boreholes are required to extract raw formation water while air is being injected in order to meet a reasonable construction time without exceeding the allowable borehole pressure.

Based on the conclusions of relevant domestic and international studies and field tests on CAES projects for the conversion of DGR, the following points are proposed:

1. (1)

   The domestic prospects for the development of CAES in DGR are promising. However, there is a need to enhance the localization of key equipment for CAES. This equipment primarily consists of compressors, heat exchangers, and expanders. While some related equipment is available in the industry, it currently holds a small market share globally, and there is still a performance gap compared to major manufacturers.

2. (2)

   At the CAES site selection stage, it is possible to drill sufficient core samples for surface testing to acquire data on permeability, mineral types, and potential sediment distribution. For testing purposes, injection and extraction wells can be positioned at different locations within the reservoir to analyze the composition and chemical properties of the extracted gas and water. This helps prevent issues such as scaling, corrosion, and oxidation, thereby reducing the impact on the porosity and permeability of the reservoir.

3. (3)

   The nature of the reservoir, geologic safety, history of depleted reservoirs, and economics need to be evaluated in advance of the project. A comparative analysis of the feasibility of the block can be made based on PG&E's evaluation of the CAES project in depleted reservoirs in KingIsland, USA, as compared to other commercial energy storage technologies.

4. (4)

   The dynamic sealing evolution mechanism of the reservoir is more complex. The numerical simulation studies of gas storage reservoir primarily focus on the reservoir itself, and there is less emphasis on the numerical simulation of the geomechanics of the cap layer. Therefore, it is necessary to combine the analysis of both the reservoir and the cap layer to conduct dynamic mechanical analysis and enhance system stability.

5. (5)

   High-flow compressed gas storage energy injection and production technology is a key technology for improving work efficiency. There are fewer studies on the safety evaluation and program of high-frequency injection and production pipe column, and further research in this area is needed to strengthen this aspect.

Based on the current situation and development of CAES technology in China, the characteristics of CAES siting in DGR, the evolution mechanism of reservoir dynamic sealing, and the high-flow pressurized gas storage and injection technology are analyzed.

The current research status of capillary confinement mechanism under alternating load, change rule of cap permeability, mechanical confinement mechanism of cap, as well as the risk of tensile, shear, fatigue and fault shear-slip damage under alternating load are summarized. Taking the project of 300 MW-10 h CAES power plant in the depleted gas reservoir of Golden Island, San Joaquin County, California, USA, as an example, this paper analyzed the safety evaluation of the pipe columns and the study of efficient injection and extraction schemes, which can provide a reference for the study of CAES engineering in DGR.