A holistic review on hydraulic fracturing stimulation laboratory experiments and their transition to enhanced geothermal system field research and operations

3.1. Conducted lab experiments based on rock type

Due to the wide application of fracturing techniques, especially in oil and gas production, many studies were performed on different types of reservoir rocks encountered to create a reliable knowledge foundation for this practice around the globe. Experiments have been performed on a variety of different rocks as named before, each analyzing not only the change of variables specific to the rock type but also general properties that are equally important in all.

For instance, for consolidated sands, parameters such as grain size distribution as well as fracture propagation across cylindrical, homogenous samples were mostly investigated under ambient and biaxial stress state. For cement, the only available study, published by Beugelsdijk et al. [18], focused on the influence of the flow rate and fracturing fluid viscosity on hydraulic fracture propagation. For concrete, an artificially constructed material, triaxial stress resistance, and the homogeneity of the samples despite different compositions were a key interest. A similar approach applies to hydrostone as the strongest and hardest gypsum cement available. For carbonates, which compose mainly of CaCO₃, fracture propagation along with stylolite structures and calcite veins are predominantly of interest and for shale, fine-grained clastic sediment, the propagation of fracture under different stress regimes is of critical importance.

Additionally, a great amount of effort has been given to sandstone as they build the basis for some of the biggest gas and oil reservoirs in the world [19]. Besides, a significant number of geothermal power plants use sandstone reservoir layers for fluid production and injection. Despite their generally acceptable permeability, in tight formations or formations with interlayer bedding, hydraulic fracturing is applied to improve the productivity of the formation. On the one hand laboratory tests have been performed for the performance analysis of the acoustic emission fracture propagation technique and homogenous sandstone samples have been fractured to investigate the fracture initiation [12], growth, and performance regarding special procedures, like cyclic fluid injection or the injection of supercritical carbon dioxide. On the other hand, the fracture propagation across natural and artificial discontinuities or even alternating layers of low permeable layers has been tested.

Aside from the individually observed parameters mentioned above, observation of the state or changes of other properties such as the type of fracturing fluid as well and formation breakdown pressure concerning minimum horizontal stress, vertical stress, or their ratio is a desirable outcome for each rock type experiment. Tables 1 to 6 provide a summary of experiments conducted on each rock type.

In Fig. 6, Fig. 7, Fig. 8, and Fig. 9, the formation breakdown pressure vs. minimum horizontal stress for different fluids and rock types are depicted.

In Fig. 10 and Fig. 11, the effect of stress ratio is drawn for concrete and hydrostone.

Fig. 12 and Fig. 13 compare the breakdown pressure vs. minimum horizontal stress concerning rock composition.

Fig. 6Breakdown pressure – minimum horizontal stress relationship – fluid types – sand

Fig. 7Breakdown pressure – minimum horizontal stress relationship – fluid types – concrete

Fig. 8Breakdown pressure – minimum horizontal stress relationship – fluid types – sandstone

Fig. 9Breakdown pressure – minimum horizontal stress relationship – fluid types – carbonates

Fig. 10Breakdown pressure – stress ratio – concrete

Fig. 11Breakdown pressure –stress ratio – hydrostone

Fig. 12Breakdown pressure – minimum horizontal stress relationship – sandstone composition

Fig. 13Breakdown pressure – minimum horizontal stress relationship – carbonates composition

3.2. Igneous rocks

Due to the significance of igneous rocks in enhanced geothermal systems, more thorough research was conducted on the previous attempts at characterizing fractures in such rocks. Granite, as the most commonly encountered rock in EGS, is an igneous rock, composing of feldspar, quartz, and glimmer, and can be produced at every stage of a Wilson cycle [57]. As geologically young igneous formations are associated with magmatic processes, depending on the formation age, their temperature is typically higher than those of other formations in equal depth. As a result, igneous rock formations are ideal candidates for geothermal energy recovery. The recovery of geothermal energy from these formations requires hydraulic stimulation, as the natural permeability of the igneous rock is too low, to allow effective fluid flow. Laboratory tests have been performed in the past with several objectives to understand the hydraulic fracturing behavior in igneous rock samples. The target was to develop an effective thermal and hydraulic fracturing scheme. Table 7 provides a detailed summary of the fluid types tested.

Tests on igneous rocks were performed under ambient, uniaxial, biaxial, and triaxial conditions. A test setup for the evaluation of the effect of hydraulic pressure on flaws, or existing fractures, under uniaxial load, has been developed at the MIT [74]. The test setup uses a transparent window to record the fracture propagation by a high-speed camera where granite and marble rock fracturing tests have been performed. First fracturing laboratory tests in 1980 indicated that rapid fluid injection rates result in tensile fractures, whereas slow injection rates cause shear failures at elevated differential stress levels [75]. There a triaxial hydraulic fracturing test system was used in the laboratory to investigate the planar and three-dimensional geometry of hydraulic fractures in homogenous volcanic rocks (rhyolite, tuff). CT scan technology was used to scan the rock samples before and after the fracturing experiment to evaluate the hydraulic fractures in three dimensions [67]. The images indicate that the hydraulic fractures cross natural pores and microcracks. No fracture propagation has been seen in zones where the natural pores are abundant since the high fracture pressure gradient of volcanic rocks generally required high fracturing pressures. The rock mechanical properties near the borehole were seen as the key factors controlling the initiation and propagation of hydraulic fractures. Additionally, a study of the reservoir stimulation process on a laboratory scale was performed by Hu et al. [64] to improve the understanding of the fracture geometry for the fracturing of magmatic rocks under triaxial conditions.

Moreover, acoustic emission monitoring was used to investigate the fracture propagation in several Sierra White granite rock samples with one injection well and four producing wells [65]. The low viscosity of the selected fracturing fluid allowed a rapid fracture development. In just 1.4 seconds, the fracture propagated near the rock’s surface. A good agreement between the reconstructed 3D fracture geometry and the acoustic emission monitoring results was found. The SEM analysis showed that the average fracture aperture is 139 μm and that the fracture follows the weak boundaries between the Quartz and Albite grains or the weak beddings in the grains.

Furthermore, Ishida, T., et al, have performed triaxial fracturing experiments with water, oil, and supercritical CO₂ [66], [76], [77]. The laboratory tests were performed on six granite samples to investigate how the viscosity of fracturing fluid affects fracture propagation. The wellbore was situated parallel to the rift plane of the granite samples. The results indicated that higher viscous fluids yield a smoother fracture pattern. In general, hydraulic fracturing induces several branch pathways around the main fracture. Supercritical CO₂ caused the most branches along the main fracture and confirms that low viscous fluids will form a more complex fracture network in rocks. The acoustic emission monitoring system showed that shear fracturing is predominant for the low-viscosity supercritical CO₂ injection, which is conformed by the microscopic analysis.

Chen et al. [24] investigated the influence of the fracturing fluid type (water, oil, and supercritical CO₂) on the fracture propagation in anisotropic granite under true triaxial load too. The evaluation used a microscopical analysis via the fluorescent method. The observed results confirm the findings of Ishida et al. [77]. Fig. 14 shows that most of the laboratory tests were performed with pure water. Supercritical CO₂ resulted in low breakdown pressures, depending on the structures of the rock sample.

Fig. 14Breakdown pressure – minimum horizontal stress relationship – fluid types – igneous rocks

Dense microcracking near coalesced fractures is an important process in altering the mechanical properties of the nearby virgin material. A detailed acoustic emission monitoring and analysis study has been performed to get insight into the formation of microcracks and their characteristics [63]. The post-processed results showed that there is a relationship between the microcrack displacement vector orientation and principal stress directions. It was concluded that tensile event displacements tended to form at different angles than shear-dominated failures.

Elsewhere, the development of microcracks in hot concrete and granite samples without confining pressure by the injection of room temperature water was investigated to develop an understanding of the thermal fracturing behavior [78]. Visual inspection, bubble leakages, pressure decay, acoustic signatures before and after the experiments, and CT scans were used to characterize the microcrack propagation. The borehole pressure decay profiles, obtained before and after each stage of stimulation indicate an increase in sample permeability.

In another effort, an experimental examination of thermal and hydraulic fracture propagation in acryl, cement, and granite samples was performed to validate theoretical results and study the fracture morphologies under uniaxial stress conditions [59]. Different sequences of thermal and hydraulic fracturing were selected. In general, the fractures were created with the faces perpendicular to the maximum principal stress. The sample breakdown pressure was almost half in acrylic specimens when thermal fracturing was applied before. Besides, the Kaiser effect [79] can be used during the thermal stimulation as an indicator of pre-existing fractures or previous stress states. The CT scans showed that thermal fractures create significant permeability, but the fractures remain almost completely closed when the temperature gradients disappear.

True triaxial hydraulic fracturing tests were as part of another study to evaluate the effect of the injection rate, the rock temperature, and the applied stress on the formation breakdown pressure of selected granite samples [58]. Visual and microscopic evaluation methods have been used. In the first test series, the injection rate was increased from initially 5 ml/min to 30 ml/min in 5 ml/min steps at a sample temperature of 150 °C. The results indicated a clear increase in breakdown pressure with the rate increase. The second test series investigated the effect of temperature. The temperature of the rock sample was increased from 30 °C to 120 °C in 30 °C steps. As high temperatures cause thermal stresses inside the rock matrix during the injection of cold fluid, the small thermal-induced cracks weaken the boundaries between grains and the breakdown pressure linearly descends with rock temperature at constant stresses. Finally, the effect of confining pressure on breakdown pressure showed that an increase in confining pressure results in an increase in breakdown pressure. The discrete elements numerical simulation showed an acceptable agreement with the experimental results.

In a different investigation, a comparison of formation breakdown pressure in granite samples at 20 °C and 120 °C was performed for a low and a high-stress state [71]. It was concluded that different temperatures do not cause a significant change in the fracture geometries, but at 120 °C, the injection pressures trend shows two peaks, and fracture pressures are increased by 3.6 to 4.9 MPa, which is controversial to the findings of Cheng et al. [80]. At high-temperature, reduced fracture energy as well as a reduction in fracture width was recorded, which indicates that high temperatures decrease the effective stimulated volume of hydraulic fracturing.

In addition, a hydraulic fracturing process in an enhanced geothermal system was simulated by true triaxial laboratory tests, applying temperatures of up to 250 °C [72]. Granite and marble samples with and without natural fractures were taken from an outcrop in a Chinese EGS project. The tests showed that the matrix of the selected hot dry rock sample without natural fractures was difficult to fracture. Natural fractures have a significant influence on fracture initiation and propagation. An extreme decrease in fracture pressure has been seen for the comparison of rock samples with and without natural fractures. A higher fracturing fluid pump rate resulted in the connection of more natural fractures. A finite element method-based numerical simulation model coupled with thermal conductivity, fluid flow, and rock mechanics was used to simulate the fracturing process.

Another modeling example is the discrete element model of naturally fractured rock presented by Zang et al. [81] as a means to simulate the seismic events of a soft stimulation method with cyclic reservoir treatment. The simulation results show a good fit for energy released by seismic events. Cyclic injection reduces the number of induced events as well as the magnitudes. Triaxial laboratory experiments confirmed the finding of the cyclic simulation treatment of smaller damage volume and more persistent fracture growth [82], [83].

Laboratory experiments on the damage evolution during cyclic injection using acoustic emission monitoring were also carried out on cylindrical granite samples [61]. The average breakdown pressure, obtained from the continuous injection experiments was taken as a reference. The cyclic injection experiments resulted in the conclusion that the level of maximum injection pressure is inverse to the number of cycles to failure and the maximum amplitude of the acoustic emission reduction during the cyclic injection. On the one hand, hydraulic energy was higher for cyclic injection, but on the other hand, the radiated seismic energy was lower.

Furthermore, the effect of six different injection schemes on the fracturing behavior of granite under triaxial load was investigated by Zhuang et al. [73]. Three schemes were injection rate controlled: constant rate continuous injection, stepwise rate continuous injection, and cyclic progressive injection, whereas the other three schemes were pressurization rate controlled: stepwise pressurization, stepwise pulse pressurization, and cyclic pulse pressurization. The fracture propagation was monitored by an acoustic emission system. The results indicated that the intragranular fractures splitting microcline, orthoclase, and quartz grains dominated the hydraulic fractures independent of the injection scheme, which results in tensile fractures. Stepwise pulse pressurization created the largest fracture length and the highest injectivity enhancement. Cyclic pulse pressurization has been known to be the best method regarding hydraulic performance improvement and induced seismicity reduction. Cyclic progressive injection generated the lowest average magnitude of the maximum acoustic emission amplitude.

Other efforts led to an investigation of hydraulic fracturing of igneous rocks under triaxial stress on a laboratory scale was performed by Siebert et al [69], [70]. Granite rock samples were tested and an acoustic emission monitoring system was used for fracture propagation monitoring. An injection procedure was developed, which reduces the negative influence of decompression on the fracture propagation by at least two injection cycles. The first cycle is applied for pre-damaging the samples by rising the injection pressure until fracture initiation. For the second cycle, the injection rate was reduced and the fracture propagated effectively into the rock.

A different work was described as testing of the mechanical impulse hydraulic fracturing method on samples of acrylic, concrete, granite, and limestone was performed under triaxial load and elevated temperature level [62], where fracturing fluids water, brine, oil, and epoxies were selected.

A benchmark test for verifying existing or new hydraulic stimulation codes was also performed based on a laboratory fracturing test, using igneous and metamorphic rock samples [60], [84]. Fracture growth and propagation were monitored by acoustic emissions and visually by ink injection. A better match between experiment and simulation results is obtained for larger values of fracture toughness, which is a result of the fact that with increasing fracture size, the fracturing behavior shifts from ductile towards brittle behavior. The final fracture extent of simulation and results are equal. The chemical response of fractured crystalline rocks to variations in temperature and stress is significant for fluid transport through fracture pathways.

In the meantime, laboratory research is going on at the U.S. DOE’s Frontier Observatory for Research in Geothermal Energy (FORGE) to validate the reactive transport and/or mechanical models for better prediction of fracture conductivity [85]. The most important parameters are identified to be the injection rate, fluid infiltration, fluid viscosity, borehole size, and injection scheme [86]. Typically, acoustic emissions monitoring is performed to detect fracture propagation in the rock samples. After fracturing, the rock samples are often analyzed by using CT scans or microscopes. Fig. 15 shows that the influence of the internal structure of the rock sample influences the breakdown pressure at low minimum horizontal stresses. Nevertheless, for large minimum horizontal and vertical stresses, the breakdown pressure is mostly dependent on their magnitudes.

Fig. 15Breakdown pressure – minimum horizontal stress relationship – vertical stress – igneous rocks

3.3. Tests combining samples of different rock types

Brenne et al. [45] performed hydraulic fracturing tests on six triaxially loaded rock samples of different compositions. Distilled water was pumped into the borehole, positioned in the center of the cylindrical sample, or into a polymer tube inside the borehole. The objective was to investigate the different fracture propagation of hydraulic and sleeve fracturing. Stable fracture propagation was achieved over a wide range of injection pressure for sleeve fracturing. The experiments showed that fracture initiation can be confidently linked to the acoustic emission count rates. A comparison to fracture modeling showed that simple fracture mechanics models can predict the experimental outcome.

A series of laboratory fracturing tests were also performed on granite, sandstone, and shale with different bedding plane directions (0, 30, 45, 60, 90), where the interlayer bonding force is relatively low [68]. Cylindrical samples were prepared and an X-ray computer tomography scanner was employed to obtain the spatial distribution of hydraulic fractures. One of the most significant parameters on the propagation of hydraulic fractures is the initial heterogeneities in rock materials. In shale, the fracture direction is diverted into the bedding plane direction while fracture extension in sandstone, shale, and granite cores differ extremely. The fractures in fractured granite cores are more tortuous than those in fractured sandstone and shale cores, which might be the result of the initial heterogeneity of granite.

Elsewhere, testing of unconventional rocks (concrete, granite, limestone), having low natural permeability and heterogeneous structures, was performed to test the mechanical impulse hydraulic fracturing method, measurement of critical state hydraulic fracture aperture, and investigate the grain-scale effects on 3D hydraulic fracture geometry [62]. Tap water, oil, and different types of epoxy were used as fracturing fluid. The laboratory results allowed insight into the hydraulic fracturing process benefiting from the direct comparison of pressure, flow rate, acoustic emissions (AE), strain, temperature, self-potential (SP), and fracture geometry data.

4.1. kISMET – USA

An interesting field test facility is the “kISMET” (permeability (k) and Induced Seismicity Management for Energy Technologies) project as a collaboration of multiple laboratories and universities, constructed in 2015 in a deep mine. The purpose of the project was to conduct hydraulic fracturing experiments in crystalline rock to understand the effects of stress field and rock fabric on fracturing efficiency and permeability enhancement, as well as to gain experience in geophysical monitoring methods. Such an environment provides the best possible stress field characterization of the stimulation operations downhole and offers close-by monitoring opportunities. The test site was about 1,500 m underground at the Sanford Underground Research Facility (SURF) in Lead, South Dakota. Five wellbores were drilled almost vertically and cored, every 3 m spaced from the next, with one injection borehole approximately in the center and a depth of 100 m, surrounded by the other 4 wellbores used for monitoring and drilled to a depth of 50 m. The project offered a great opportunity to gain experimental experience. Additionally, it was observed that hydraulic fractures were formed quite uniformly, meaning that the rock fabric did not have a significant effect on controlling the fracture orientation. Seismic monitoring also only recorded a marginally higher signal when the deepest fractures were created [87].

4.2. Collab – USA

The US Department of Energy-funded research to be carried over several years by several partners and laboratories, referred to as the EGS Collab project. The main aim of the project is to improve and validate the thermal, chemical, mechanical, and hydrological models with the data obtained from the field. The tests are carried out at the 1500 m deep SURF in South Dakota, at intermediate scale test beds (~10 m). The test bed is consisted of eight wellbores, having a depth of 60 m, six used for instrumentation and monitoring, and the other two are used for injection and production. The results are used to support the FORGE project. At the site, three experiments were planned to be conducted:

- Experiment 1: To perform and characterize several heavily monitored stimulations. Measuring stimulation behavior, permeability enhancement, and rock characteristics would provide insight into the nature of the stimulation (fracturing, shearing, mixed-mode or thermal) under similar-to-reservoir conditions [88].

- Experiment 2: The experiment is similar to Experiment 1, but with the difference that it focuses on hydraulic shearing of an existing fracture, as opposed to hydraulic fracturing [89].

- Experiment 3: The experiment will investigate other alternative stimulation techniques or general operational methods to better extract heat from the EGS reservoir [90].

4.3. FORGE Utah – USA

Frontier Observatory for Research in Geothermal Energy (FORGE) is driven by the U.S. Department of Energy’s program. The program came to existence to develop and practice techniques that allow for accessing, exploiting, and monitoring EGS reservoirs in different locations around the United States, to demonstrate the potentials of the EGS technologies and their role in the power generation to the energy industry, investors and the public. The FORGE program was performed in three phases. Phase 1 was the preliminary study of existing data from five different sites around the US. Eventually, in 2018, the University of Utah’s Milford, Utah site was selected to be the first site for the FORGE laboratory. As part of Phase 2, well 58-32 was drilled with a total depth of 2,290 m into permeable, crystalline rocks to provide a direct measurement of reservoir parameters such as temperature, rock type, permeability, and stress in the reservoir, as well as to define the groundwater depth. In phase 3, two additional wells were drilled and completed to be used for seismic monitoring and to complete the infrastructure for research and development, testing, and observation [91]-[93].

4.4. GTS – Switzerland

In Switzerland, as part of the government energy strategy 2050, the Swiss Competence Centres for Energy Research (SCCER) has initiated a research network to conduct innovative and sustainable research on renewable and emission-free forms of energy. To investigate the geothermal potential, two test sites were selected in the Swiss Alps where in-situ tests could be performed under conditions that are as close as possible to an EGS.

The first site, called Grimsel Test Site (GTS) was used for experiments from 2015 and 2017, on the scale of decametre. The second site called the Bedretto Underground Laboratory for Geoenergies (BULG) will be compatible with tests performed on the hundred-meter scale and will be explained later in the paper. The significance of these test sites is in the fact that they are comparable in terms of scale to the Basel EGS project. These in-situ tests will be supported by laboratory tests as well. The GTS was initially developed in the 1980s for the disposal of nuclear waste and is located 450 m below the surface within Crystalline rocks, in particular granodiorites. Both, hydro-shearing and hydro fracturing, was tested at the site in six experiments conducted on two wellbores with 45 m depth and water are stimulation fluid [94].

5.1. Desert peak – USA

Well 27-15 of the Desert Peak geothermal field in Nevada was a proper candidate for fracturing as its performance as an injector or a producer had been dissatisfactory from the beginning. Nevertheless, due to its desirable downhole temperatures (up to 190 °C) and the existing surface facilities, fracturing operations were executed to reach several goals, including the development of a site-specific fracturing plan showcasing the most practical and economical techniques, as well as investigating the possible merits of power generation from the entire EGS project. Supported by the Department of Energy, the stimulation plan, conducted between 2010 and 2011, consisting of multiple stages, including multiple low-pressure shear stimulation phases, two stages of chemical stimulation. Finally, an extensive phase of hydraulic fracturing at high pressures concluded the program. The target was to investigate whether or not the injectivity of well 27-15 reaches the commercially accepted limit in the Desert Peak. This target was achieved and as a result of the stimulation, the injectivity of well 27-15 increased by a factor of 60 and confirmed via the tracer test. It was also observed that the main shear stimulation success was in the early stages of the program, suggesting that a shorter program at higher injection pressures could potentially offer similar results. Additionally, it was observed that chemical stimulation tends to have better efficiency when used solely to improve an existing network of fractures, meaning that using chemicals for fracture initiation can lead to accumulation of the fluid in the near-wellbore region and as a result costly operations [96].

5.2. Raft river – USA

Another successful fracturing job was performed in Raft River, Idaho, as part of the demonstration projects of the Department of Energy. The Raft River site includes four injectors and four producers, and the well RRG-9 ST1 as the newest injector was stimulated using a combination of hydraulic and thermal methods. The hydraulic stimulation plan began in 2012, followed by a year and a half of shut-in period to construct proper surface pipelines connecting to the wellhead. The well was then stimulated twice in the following years with periods of shut-in and regular cold water injection. By the end of the fracturing operation, the result of the modified Hall plot as well as the increase in injectivity confirmed the success of the fracturing job [97], [98].

5.3. Newberry – USA

The Newberry EGS demonstration project began in 2010 and was funded by the Department of Energy in the USA. Located in the area of a volcano in central Oregon, the well NWG 55-29 was originally drilled in 2008 to a depth of 3,067 m, encountering temperatures of about 331 °C. After the approval for the funding and 2 years of development and planning, the well underwent the first round of stimulation in 2012, which led to damages of the casing. After installing a tie-back over the damaged zone and conducting further remediations, the well was stimulated once again in 2014 in the form of hydraulic shearing. The overall outcome was an increase in the permeability as well as the volume of fractures in the microcrystalline granodiorite reservoir, with a five-fold injectivity increase. In this stimulation process, a special component called TZIM (Thermally-degradable Zonal Isolation Material) was used. The material is made of components that degrade without any environmental harm within a certain temperature range. After implementation, it was evident that two thief zones were blocked and one or two new zones were opened as a result of using TZIM. Seismic monitoring was performed continuously to provide a clear understanding of injection limits. The partner well 55A-29 was drilled in 2014 to act as the production well of the doublet [99].

5.4. Rittershofen – France

The ECOGI is the first EGS project in France to produce water from geothermal reservoirs in the Upper Rhine Graben. The project location is in Rittershofen, about 6 km east of the well-known Soultz-sous-Forets EGS. Initiated in 2004, the feasibility study concluded in 2011 and the first well GRT-1 was drilled between September and December 2012 to a depth of 2,580 m MD into a granite formation. The second well GRT-2 was then drilled between May and July 2014 to a depth of 3,196 m MD to be the producer. Stimulation in GRT-1 was conducted in three phases: 1) a thermal stimulation, 2) a chemical stimulation followed by 3) a hydraulic stimulation. The environmentally friendly acids based on Glutamate were used in phase two to dissolve the minerals that seal the natural fractures. The job was deemed successful and the wells were later put into operation [100].

5.5. Otaniemi – Finland

Otaniemi is the geothermal doublet project developed in Espoo, Finland, in the vicinity of Helsinki for district heating purposes. Being the deepest EGS project to date, the main wellbore OTN-3 was drilled to a depth of 6.4 km in 2018, reaching temperatures of about 120 °C. A 10 m offset well, OTN-2, was also drilled to 3.3 km for monitoring purposes, which was planned to also be extended in 2019 to 6.1 km. The wells are almost entirely in a granite and pegmatite formation and have therefore been drilled using air and water hammer drilling mechanisms, together with rotary systems for better steering. OTN-3 was hydraulically fractured later in 2018 for 49 days and in five phases each with different injection pressures. Since the project is located underneath a largely populated area, even a slight chance of seismic activity can not only be dangerous but also affect the positive image of the project in the public’s eyes. Therefore, a TLS (Traffic Light System) was developed for the monitoring of seismic activities that could assist the mitigation and define the steps to be taken under various circumstances. The system can define the limits of pumping power, pressure as well as the shut-in time between the different stages of stimulation [101]-[103].

5.6. DEEPEGS – Iceland

The Iceland Deep Drilling Project (IDDP) was kicked off in 2000. After years of study and planning, the first drilling project was meant to start in Rejkyanes geothermal field located in the southwest of Iceland in 2005. Yet, the deepening of some existing wells hit several issues, persuading the team to choose another location in the northeast of Iceland for drilling the IDDP-1. The well could not reach supercritical conditions as it hit a magma layer at only 2,100 m, with fluid pressures and temperatures lower than expected. After two successful years of operation, and for maintenance reasons, the well had to be quenched, which led to its collapse and eventually its abandonment [105]. The exploration well IDDP-2 was therefore drilled and completed in 2017 as part of the DEEPEGS demonstration project conducted by Iceland and France and funded by the EU Horizon 2020. The final depth of the well was 4,650 m, and although expected to reach supercritical temperatures up to 600 °C, the recorded temperature after six days of heating at 4,550 m was 426 °C. Among the problems faced during drilling was a complete lost circulation at depths below 3.2 km despite which the drilling continued. The formation rock at that depth consisted of micro-dolerites with basalt. During the drilling process and for over a year after, cold water was injected not only to cool down the equipment but also to enhance the permeability in the near-wellbore region through thermal stimulation. The total volume of injected water was therefore in the range of 1.5-2 million tons, reaching a final injectivity value of 2.7-2.9 l/s/bar [105], [106].

5.7. Pohang – South Korea

The project in Pohang was the first EGS developed in South Korea and was kicked off in 2010. Wellbore PX-1 with 4,127 m depth and PX-2 with 4,348 m depth were drilled to be stimulated and used together with a third wellbore to create a triplet system. The reservoir was made of granodiorite and granitic gneiss with temperatures reaching 140 °C. PX-2 was the first well to be stimulated in 2016. After three stages of hydraulic fracturing lasting until 2017, the injectivity of the reservoir increased by 2.7 times. Shortly after the start of fracturing in PX-2, hydraulic shearing was conducted in PX-1. Stimulation in this well also led an almost twice the injectivity index [107], [108]. The further development was put to rest after a 5.4 magnitude earthquake happened in the location in November 2017, two months after the last stimulation took place. This was the second-largest earthquake to have ever happened in South Korea. It caused injuries to 90 people and the estimated property damage was about $52 million. A further study investigated the relationship between seismic activities after each injection. Based on that, and the lack of any other activity before the operation of the EGS, it was proven that the stimulation operations were the reason behind the earthquake [109], [110].

Fig. 18 shows the increase in breakdown pressure with minimum horizontal stress for the field cases and confirms the finding of the laboratory tests. Typically, the minimum horizontal stress is in the range of 20 to 40 MPa, resulting in formation breakdown pressures of 20 to 40 MPa too.

The trend of the increase in formation breakdown pressure by the increase in injection rate, also for field cases, is shown in Fig. 19. Nevertheless, other factors, like natural internal rock structure, fluid type, etc. that influence the formation breakdown pressure were not found and therefore no general statement can be made.

6.1. Schneeberg – Germany

In Saxony (Germany), a project is aimed to be carried out in crystalline rocks, in particular granite and phyllite. A roadmap has been created by 2015 in a collaboration between multiple companies and universities, with the support of the German Federal Ministry. The initial goal is to drill a wellbore “Schneeberg 1” to acquire geological and reservoir data. The wellbore is expected to be drilled to 5250 m MD (5000 m TVD), crossing the fault system Rote Kamm at 2000 m. Temperature experienced at this depth would be between 180 °C and 200 °C. The borehole should be capable of producing 30 gallons/s [111].

6.2. SHEGS – Hungary

The South Hungarian EGS (SHEGS) is a demonstration project in south-eastern Hungary, funded by the Hungarian Ministry of Economy. The project includes four production wells with an approximate depth of 4 km each, and two re-injection wells, all connected to a power plant with an estimated capacity of 8.9 MWe. The reservoir is composed of granitoid rocks and the overall thermal gradient ranges between 55-65 °C/km. Seismic activities are planned to be closely monitored to avoid any issues for the population as well as the available infrastructure [112]. The drilling was aimed to start in 2016. However, the project is still not in operation and no further information was provided.

6.3. Qiabuqia – China

Despite the existence of vast geothermal resources in China, they only recently started to gain more popularity. These resources are estimated to be equivalent to 860 million tons of standard coal, which can support China’s energy needs for the next 3,945 years even at a 2 % exploitation rate. As a result, the first demonstration project in China was established in Qiabuqia geothermal field in the northwest of China. Ten exploration wellbores were drilled on the site, with many reaching temperatures as high as 178 °C. Well GR1 was, particularly of interest. It was drilled to 3,705 m in 2017, reaching a stable temperature of 236 °C after five months of a shut-in. Nevertheless, based on the stress state of the surrounding granite formation and the existence of natural fractures, the pay zone was selected to be between 3,200 and 3,705 m, with a temperature of 218 °C. To achieve better flow rates, multiple stimulation scenarios were modeled, concluding that hydraulic fracturing is the best approach for future planning [113].

6.4. DEEPEGS – France

The DEEPEGS project was conducted as a collaboration between Iceland and France. While the project aimed for the development of high-enthalpy EGS with supercritical pressure and temperature in Iceland, it aimed to explore the geothermal capacities in hydrothermal reservoirs in southern France in Valence and Vistrenque. Due to lack of public acceptance in Valence, a secondary location was chosen in Riom, in central France. Yet, the planning and acquisition of permits hit a dead end, and eventually, the project team decided to replace the two target projects with a geothermally rich site in Vendenheim [114]. The second demonstrator wells in Vendenheim were therefore drilled in 2019 as a doublet and tested in 2020. The wells reach the Upper-Rhine-Graben at depths of 5,308 m and 5,393 m and temperatures of 200 °C. Since the measured injectivity of the zone is below the intended value, thermal stimulations are considered for future steps [105].

6.5. BULG - Switzerland

The BULG is an ETH facility in a 100 m long cavern at depths between 1,000 to 1,200 m, with temperatures reaching a steady value of 17 °C. Site characterization was conducted in 2018 and experiments are expected to begin in 2020 with several project partners. The rock type is the particular Rotondo Granite. The depth, location, and stress regime of this test site are superior to GTS in terms of similarity to an actual EGS project [94], [115].

7.1. United Downs – UK

The United Downs project is the first geothermal power project in the UK. It is located in the west of Cornwall and is consisted of a production well with a depth of 4,500 m and an injection well with a depth of 2,500 m. The formation rock is mainly granitic and has an approximate temperature of 190 C. The aim is to generate 1-3 MWe electricity from this field. Drilling and completion of the wells were planned to be completed in 2019. Since the reservoir is naturally fractured, with a shearing onset of 5 MPa, no fracking procedure is foreseen at the moment but could become necessary in the years to come. It is expected that a normal injection of water can support the pressure requirements. The success of this project can pave the way for future geothermal projects in the area and fulfill the goal of generating the country’s electricity from renewable sources by 2030 [116].

7.2. Eden – UK

Eden geothermal demonstration project aims to recover energy from the naturally fractured granite formation beneath the city of Cornwall. Initially, production will be through a coaxial circulation in one well with an approximate depth of 4.5 km. The energy is supposed to be captured and used for heating greenhouses and offices in the area. In phase two of the project, a second well will be drilled to the same depth, converting the system into a geothermal doublet and connecting it to an electric power plant. The first well EG1 will be drilled in Spring 2021 after some delays, using conventional rotary drilling. Since the reservoir is naturally fractured, no stimulation is foreseen at the moment but could be a possibility subject to future production trends [117].

7.3. Mauerstetten – Germany

To investigate the geothermal potential in the Western Bavarian Molasse Basin, a geothermal well was drilled together with a side-track close to a fault system. The main well, GT1 was drilled to 4,523 m MD (4,085 m TVD) and the side-track GT1a to 4,052 m MD (3572 TVD) with a lateral distance of 513 m to GT1 in 289-degree azimuth. Because porosity generation by recrystallization is a slow process, natural fractures are considered the main network for fluid flow. Nevertheless, the calcite fillings of such fractures had reduced the transmissivity of the fractures. Therefore, an acid stimulation plan with HCl pills was already conducted. Based on the results, a future stimulation plan can be prepared, however, no further data is available at this point [118].