Tectonic Collision, Orogeny and Geothermal Resources in Taiwan

Abstract

The recent tectonic evolution of Taiwan created ideal conditions for geothermal resources: heat, water and permeability. We examine heat flow measurements, seismic tomography, seismicity, hot spring distribution, tectonic history, geology, and volcanism described in previous studies to understand the relation between tectonics and geothermal potential in Taiwan. Taiwan is the youngest tectonically created island on earth. The island formed as a result of the transition from subduction of the Eurasian Plate under the Philippine Sea plate to active collision. Collision results in orogenic mountain building. The geology of the island is primarily an accretionary prism from the historic subduction. This active orogeny creates unusually high geothermal gradients by exhumation of the warmer material from depth and by strain heating. As a result, temperatures reach up to ~200 degree C. Volcanoes in the northern tip of Taiwan provide an additional source of heat. Favorable fluid flow from meteoric waters and permeability from seismicity and faulting results in exploitable geothermal systems near the surface. These systems can potentially provide geothermal power generation throughout the whole island, although there are currently only two geothermal power plants in Taiwan.

1. Introduction

Most of the energy consumed in Taiwan is produced by imported fossil fuels. The development of renewable energy is a priority, both to move the country toward energy independence and to combat global climate change. Power produced from geothermal energy is an important component of a renewable energy portfolio because it is continuously available, as opposed to wind and solar, which are intermittent in availability.

The presence of numerous hot springs and high measured heat flow throughout Taiwan are indications that significant geothermal energy resources are present, even though only two geothermal power plants are currently operating in the country. Figure 1 shows the distribution of hot springs. The hot springs and high heat flow at the northern tip of Taiwan are associated with known volcanoes, but, interestingly, thousands of hot springs are also found in many other areas from the north to the south where no volcanism is found. The potential for geothermal energy from these sources is perhaps more relevant for Taiwan than the limited volcanic area, as their spatial distribution is much wider and thus can perhaps serve a much larger area.

Figure 1.

Hot springs of Taiwan. Light beige indicates the central range, with foothills to the east and the Longitudinal Valley (a long green strip) east of the foothills. From [1, 2] the blue box identifies the area of the highest heat flow in Taiwan.

In this chapter we examine previous studies to show the relation between evolutionary tectonics and the potential for geothermal resources in Taiwan. We show how this evolution created ideal conditions for geothermal resources; heat, water and permeability. We examine heat flow measurements, seismic tomography, seismicity, hot spring distribution, tectonic history, geology, and volcanism described in previous studies to understand the relation between tectonics and geothermal potential in Taiwan.

2. Evolutionary tectonics

Prior to 6–9 Ma, sea floor spreading occurred in the South China Sea and the lithosphere within the latitudes of Taiwan was subducted eastwards beneath the Philippine Sea Plate along the Manila Trench [3]. Convergence was at a 5.6 cm/year with respect to Eurasia [3, 4]. The convergence created an accretionary prism of varying ages, a result of off-scraping sediments and rocks from the under-thrust plate and depositing sediments and volcanic ashes from volcanic arcs on the overriding plate [5, 6]. The sediments are a result of marine sedimentation occurring on the South China Sea oceanic crust during the Miocene [5, 7]. Taiwan is an accretionary prism from the previous subduction [8]. About 6–9 Ma sea floor spreading ceased in the South China Sea leaving the Eurasian plate subducting eastward under the Philippine Sea plate. At the same period the Luzon arc acquired a significant enough topographic expression to resist subduction and start to collide with the Eurasian plate. The sediment strata began to show evidence of plate collision in early Pliocene and rising of Taiwan, about 5 Ma [8, 9]. Active subduction continues today north and south of Taiwan (Figure 2).

Figure 2.

Major tectonic features near Taiwan (modified from [10]). Large red triangles indicate direction of subduction. Small red triangles along the north of the island are the volcanoes of the active volcanic front of the Ryukyu arc [11]. Volcanoes to the southeast are part of the Luzon arc.

The Luzon volcanic arc (LVA) is an intra-oceanic volcanic arc which belongs to the Philippine Sea plate (PSP). Today the LVA is in contact with Coastal Range of Taiwan, along the Longitudinal Valley in south western Taiwan [3, 4, 12]; Figures 2 and 3. Fission track dating in the Central Range indicates a gradual rising until about 2 Ma when it began to accelerate [2, 4] and caused the Central Range to experience rapid uplift of the roughly 40 km wide range for most of the length of Taiwan; the highest mountain uplift in the world, as much as ~26 mm/year. Other regions experiencing present-day uplift in Taiwan at rates of ∼0.2–25.8 mm/yr. include the Hsuehshan Range, the Central Range, and the southern part of the Western Foothills [14, 15] Figure 4. An uplift rate of 22.9 mm/yr. occurs in the central northern portion of the Central Range, and the southern part of the Coastal range displays uplift rates of ∼1.3–25.8 mm/yr., the largest in Taiwan. The uplift rates decrease rapidly toward the north and diminish gradually toward south [17].

Figure 3.

From [13]. Thick red lines are faults. Triangles point to subduction, plain line is the Philippine fault hypothesized to extend into Taiwan. Description of other feature are available in [13] the Taitung (1) and Leyte (2) geothermal prospects are located along the left-lateral strike-slip Philippine fault [13]. Black dots shown the location of the Luzon volcanic arc.

Figure 4.

Provinces (from [16]).

A number of studies agree oblique collision results in the evolution through time of Taiwan mountain building visible as a continuum from the present-day Manila subduction system to the south (before collision), through middle Taiwan (collision) and northeast of Taiwan, across the southern Okinawa Through and Ryukyu subduction system (post-collision) [3, 4, 10]. Also, the age of metamorphic geology increases from the west to the east across Taiwan [18].

There are many thrust faults and folded anti- and syn-clines roughly perpendicular to the convergence. Active seismicity and highly fractured zones create permeability. The 1999 M = 7.8 Chi-Chi earthquake occurred on a thrust fault in the Central Range and demonstrates one way stress buildup from collision is released [19, 20]. The Longitudinal Valley is oblique to the convergence and hosts a left-lateral strike-slip fault, which may be an extension of the Philippine Fault that runs through the Luzon and several smaller islands of the Philippines (Figure 3). The saturated fractured geology creates the pathways for hot water and steam to come up to the surface making the Taitung and Leyte geothermal prospects (Figure 3). Sibuet et al. [13] connects this fault with the Longitudinal left-lateral fault. In Taiwan, most geothermal prospects are located along this key fault.

3. Geology

Seven provinces of the Taiwan orogen can be recognized (Figure 4): (1) the Coastal Range, along the northern extension of Luzon arc, consisting of fore arc sedimentary units and andesitic volcanic rocks; (2) the Longitudinal Valley, filled with young sediments, being the plate boundary between the Eurasian and Philippine Sea Plates; (3) the eastern Central Range, consisting of pre-Tertiary (Tailuko belt) and Miocene (Yuli belt) metamorphic complex rocks; (4) the western Central or the Backbone Range (capped by Miocene slates), (5) the Hsuehshan Range (Eocene-Miocene slates); situated in the northern half of the island and tapers off toward the south, (6) the Western Foothills, the fold-and-thrust belt, composed of clastic Oligocene-Pleistocene sedimentary rocks, and; (7) the Coastal Plain, containing younger sediment deposits [16, 21]. Longitudinal Valley represents the tectonic suture zone separating metasedimentary sequences of the Central Ranges from the accreted sedimentary and volcanic arc rocks of the eastern Coastal Range.

Continental crust of the Chinese continental margin colliding with the Luzon Volcanic Arc has deformed Miocene to Quaternary sedimentary marine stratigraphy into easterly-dipping fold and thrust belts of the Western Coastal Plains, Western Foothills and the Hsuehshan Range in western Taiwan Island. To the east a metamorphosed continental margin sequence has been exhumed along westerly dipping faults as the Central Ranges. Along the eastern side of Taiwan Island, the Coastal Range represents the northern extent of the Luzon Volcanic Arc, which has been accreted onto the eastern margin of the exhumed metamorphic rocks. The geology of the accretionary prism of the eastern Central Range is primarily made up of a huge sequence of deep-sea turbidities (Miocene Lushan Formation). It consists of mudstones, siltstones, and sandstones [16, 22].

4. Heat

In Taiwan, in the mature collision zone, the heat flows are high. Heat flow in the mountains is mostly between 80 and 250 mW/M² [23]. Overall, the high heat flow in the mountainous regions is interpreted as a result of rapid uplift and exhumation of the warmer material from depth [24, 25, 26]. However, several researchers in orogenic zones have shown that uplift alone is not sufficient to account for the high heat flow [27, 28, 29, 30]. Theoretical and geological considerations suggest that viscous heating is a cumulative process that may explain the heat deficit in collision orogens; where severely deformed rocks over a short time span cause viscous heating and can account for this deficit and explain further up-warping of the isotherm [31, 32]. Whereas radiogenic heat production can be inferred from measured concentrations of radioactive elements and heat flow in stable regions of the lithosphere, the contribution to heating by deformation can potentially be measured only in actively deforming lithosphere where it may not be easily separated from other sources of heat. The strain heating and upwarping together, create favorable geothermal gradients in the Central mountains of Taiwan. The shallowest isotherm in the Central Range may also account for the aseismic zone, likely due to too hot and pliable crust unable to sustain enough stress to generate earthquakes [8]. Other sources near surface heat may include groundwater circulation, topographic effects, and higher radiogenic heat production rates in the continental crust. Complex deformations in lower crust and upper mantle following the collision might have also affected the thermal structures in this region [33].

Heat flow gives valuable insight into evaluating the tectonics of a region, and the geothermal gradient is the primary initial indicator of a viable geothermal resource. Heat flow is derived from the geothermal gradient and thermal conductivity, which are typically obtained from temperature measurements in pre-existing wells or laboratory samples. Heat flow is generally described in units of milli Watts per meter squared (mW/M²), and the geothermal gradient is described as degrees Celsius per kilometer (^oC/km).

Figure 5 shows heat flow measurements from [25]. They studied heat flow in Taiwan using previous values obtained from [27, 33] and added many values obtained off-shore of Taiwan. Chi and Reed [25] points out that there is still debate whether the heat flow data from some of the “geothermal wells” are representative of the regional heat. However, they also point out that several studies are able to fit the high heat flow pattern by thermal modeling using different crustal kinematic models [33, 34, 35]. Within Taiwan, and along the central range, heat flow reaches over 300 mW/M², whereas the worldwide average is about 50 mW/M².

Figure 5.

Heat flow measurements from [33]), re-plotted from [25].

Chi and Reed [25] identified a dramatic difference in heat flow between the subduction zone to the south, where values are near world averages, and the collision zone in Taiwan. At a latitude of ~20.5°N (not shown in Figure 5), in the subduction zone, heat flow decreases even further from about 75 to 40 mW/M² from the trench to the upper slope domain of the accretionary prism. To the east in the fore arc basin, heat flow values are ~25 mW/M² (Figure 5). The heat flow pattern along this transect is consistent with the three in situ heat flow measurements farther to the south at ~19°N [33, 35]. Heat flows in the satellite basins in the arc region are ~50 mW/M². Farther to the north in the initial collision zone (Figure 5), the continent-ocean boundary (COB) enters into the trench near 21.2°N. Hwang and Wang [36] have collected 12 thermal probe measurements along a transect from continental shelf (117°E, 22.8°N) to continental slope (118.1°E, 19.3°E) that is 220 km west of and parallel to the trench. Chi and Reed [25] treat this data set as the initial condition before the Chinese passive margin enters into this convergent boundary. Hwang and Wang [36] found that heat flows are ~80 mW/M² in the continental slope and decrease to 70 mW/M² in the abyssal plain. Chi and Reed [25] also found slightly increased heat flows once the incoming sediments were scraped off and incorporated into the toe of the accretionary prism, where intensive dewatering occurs. High geothermal gradients (40–80°C/km) and heat flows (50–105 mW/M²) were found in a thick basin near the toe in this region east of the COB (Figure 2). This might be a result of intensive dewatering in this basin, which covers a circular region with a diameter of 60 km centered at 119.8°E, 21.6°N. Seismic reflection data show conjugate fault plane reflections within the basin, even though the displacements across the faults are small, suggesting possible fluids within the fault zones. Away from the toe in the initial collision zone, the heat flows decrease toward the arc as the sediments stack thicken, especially from lower slope domain to the upper slope domain. Heat flows ranging from 30 to 60 mW/M² and increasing toward the arc are identified in the back-thrust domain.

Taiwan has a very good geothermal gradient. The average land heat flow in the world is about 30 mW/M². If there is a local heat flow value greater than the average, then there is a good potential for geothermal development [37]. Song and Liu [37] identified nine major geothermal resource regions based on anomalous geothermal gradients >35°C/km (Figure 6): (I) Tatun volcanic area in the north; (II) Yilan Plain along the Lishan Fault, it extends southwest to Mount Lu, covering Jiaoxi, Qingshui, Tuchang, Lushan and other geothermal areas; (III) For regions with abnormal geotemperature gradients higher than 35^o C/km Lushan; (IV) Ruisui-Antong; (V) Wulu-Hongye; (VI) Zhiben-Jinlun; (VII) Baolai; (Viii) Guanzailing; (IX) Hsinchu-Miao Li. These nine high anomalous areas of geothermal gradient are the potential areas of geothermal development. Taiwan is currently conducting small-scale exploration to potentially exploit these areas.

Figure 6.

a. Temperature gradient map of Taiwan showing nine EGS regions with anomalous high temperature gradients: (I) Tatun; (II) Chingshui-Tuchang; (III) Lushan; (IV) Juisui-Antung; (V) Wulu-Hungyeh; (VI) Chihpen-Chinlun; (VII) Paolai; (VIII) Kuantzuling; (IX) Hsinchu-Miaoli. b; potential Mwe of the nine zones. From [37].

5. Geothermal reservoirs

Organic geothermal reservoirs generally are permeable zones with fluid flow and heat. Taiwan has about 2500 mm of rain each year [38], sufficient for replenishing reservoirs. Engineered geothermal reservoirs only require heat, and fluid and permeability are created artificially. Heat is the important source for geothermal development in both cases. In Taiwan, orogenic reservoirs are typically located at a shallow depth of 2–3 kilometers in the Central range and volcanic zone where the geothermal gradient is greater than 30^o C/km. The development depth of enhanced geothermal away from the Central range is usually greater than 3000 m where the geothermal gradient is approximately 30 mW/M². Exploration is necessary to determine whether there is an organic reservoir or a candidate for an engineered system. Within areas outlined in Figure 6, there are 108 geothermal potential locations for geothermal development that are located at a shallow depth of 2–3 kilometers due to high geothermal gradients [39]. Except for the Tatun and Keelung Volcano Groups in the north, the main high-gradient areas are located in the metamorphic rock belts.

The Tatun volcanic group is located in the northernmost part of Taiwan, mainly composed of more than 20 Pleistocene andesite volcanoes such as Tatun Mountain. At Tatun, there is a classical lava cone, and there are many hot springs, fumaroles, sulfur pores [40, 41], and other indications of intense geothermal activity. The main shallow geothermal reservoirs are located between the Jinshan and Kanjiao faults. The temperature is about 200-290^o C [40]. Deep geothermal wells have encountered a high temperature of 293^o C, which is the highest temperature currently reached in Taiwan [25].

Organic geothermal reservoirs can be roughly divided into two types: hot water type and steam type. Engineered geothermal system (EGS) is another type of reservoir. Organic geothermal energy development is limited by hydrothermal production capacity. In geothermal fields or reservoirs, abundant and high-temperature geothermal water and well-developed fissure structures are required. The hot water type geothermal system is based on the presence of hot water in the reservoir. The water phase controls the reservoir pressure, and its temperature is the highest. Temperatures can range from less than 100^o C to 370^o C, but geothermal system above 200^o C are optimum for power generation. The vapor type of reservoir is formed by the high temperature heat source supplying heat and the low permeability of the rock formation. Reservoirs can evolve from water type to steam type if the amount of hot water extracted is more than the amount of that replenished from groundwater. In the vapor reservoir, two phases of hot water and vapor coexist, and the vapor phase controls the storage layer pressure. The hot water phase flows in small pores due to high surface tension, while the vapor phase escapes through larger conduits in the upper geothermal system. Currently, steam type reservoirs only account for 10% of global geothermal production.

Engineered geothermal system typically create permeability between two wells by hydro-fracking or controlled fracturing from injection of cold water. Figure 7 shows a scheme of an engineered system. This approach has been largely unsuccessful to date. This is because one, the fluid is not heated sufficiently in the short distance traveled, and two there has been difficulty creating a large enough fracture system, which would help in solving the first problem. Because the general gradient is about 30^o C/km, so the development depth of enhanced geothermal is usually greater than 3000 m.

Figure 7.

Engineered geothermal reservoir.

According to the geothermal research report [42], if 2% of the thermal energy stored in the 3–10 km rock layer of the earth’s crust can be obtained the heat energy produced by it is as high as 2.8 × 10 5 EJ, which is about 2800 times the total energy consumption of the United States in 2005. There is a huge geothermal potential for engineered geothermal systems deep underground in Taiwan. If it can be successfully developed, it will be an important independent energy source in Taiwan.

Two geothermal power plants, a total of 6 MWe, have started to produce electricity in Taiwan at the end of 2021. Two other small geothermal plants, 1 and 2 MWe, are being developed in indigenous areas, and are planned to operate by the end of 2022. The government has established a short-term geothermal goal to reach 200 MWe by 2025 with expansive exploration, test drilling, and power plant management. Further, the government is investing in developing engineered geothermal systems to reach long-term goals. It is hoped Taiwan may reach the GWe-level geothermal power production by 2035, thereby reducing the carbon dioxide emission as part of a global-village member.

6. Methods for exploration

Geophysical methods for exploring for orogenic geothermal reservoirs include the gravity, aeromagnetic, electro-magnetic (magnetotelluric), and tomographic imaging. The gravity method models the spatial distribution of rocks with different densities at depth to match the measurements of the acceleration of gravity at different positions on the surface. Figure 8 shows contours of gravity in mgals in southern Taiwan [43]. Geothermal reservoirs typically occur at low gravity values as seen below RF1, RFL1, RF2 and RFL2 where the gravity contours show significantly lower values than surrounding rocks.

Figure 8.

Gravity showing low density rocks with geothermal potential at RF1, RFL1, RF2 and RFL2 along faults. Square in sub-figure shows the area, which is the same as the square showing high heat flow in Figure 1.

An aeromagnetic survey records the magnetic field from an air plane flying over the area of interest. Magnetic field measurements are typically at 500 m intervals and the survey area is crisscrossed in parallel and perpendicular directions. Typically, a helicopter is flown over at a height of about 500 m above topography. In addition, magnetometers are used to continuously measure the geomagnetic field at a chosen base station on the ground. Normally, correction for the International Geomagnetic Reference Field (IGRF) involves removing it in order to show only magnetic anomalies related to geology. Resolution can be as high as 100 m from the surface to 10 km depth [44], Figure 9 shows typical results over a geothermal reservoir. In this case, the linear features of the magnetic susceptibilities (H1, H2, and H3) were interpreted as being separated by large linear vertical faults. Although there is no direct evidence of a geothermal reservoir in this particular aeromagnetic survey, the delineation of structures such as faults can aid in the search process.

Figure 9.

Magnetic susceptibilities. From [44].

Electromagnetic methods can be either passive, utilizing natural ground signals (e.g. magnetotellurics) or active, where an artificial transmitter is used either in the near field (as in ground conductivity meters) or in the far field (using remote high powered military and civil radio transmitters as in the case of VLF and RMT methods). Magnetotelluric (MT) surveys estimate the Earth’s electromagnetic impedance by measuring naturally occurring electromagnetic waves in a very broad frequency range. They typically record the full component MT data (i.e., Ex, Ey, Hx, Hy and Hz) induced by natural primary sources and measured relatively uniformly at 1000 m intervals across the area of interest. Frequency range is generally from 10 kHz down to 0.01 Hz and can be even lower when sounding duration is as long as five days. Interpretation is based upon inversion of the MT data to derive resistivity values. The final 3D model used for interpretation is the one with the lowest root mean square misfit. Figure 10 shows typical MT results at three depths in a reservoir in southern Taiwan. The low resistivity values (green) can indicate the location of a reservoir. The underground resistivity is very low resistance layer of 10 Ω-m. When the rock layer is subjected to hot water, the resistivity of the formation will be significantly reduced. Further abnormally low (below 100 Ω-m) area may indicate a heat source.

Figure 10.

Resistivity values at three depths. From [45].

A recently developed seismic method utilizes passive earthquake sources and dense recording networks to image reservoirs, and has shown promise in seismically active areas [46, 47]. Inexpensive recorders and automated data processing makes this possible in a short amount of time at a minimal cost [48]. The propagation energy from the earthquakes passes through the geology to recording systems at the surface. Tomographic inversion is used to back project these recordings to provide the images at depth [49]. The primary information comes from propagation of first arriving P- and S-waves and their pulse widths. These provide P- and S-wave velocity and Qp and Qs attenuation structure throughout the volume [46]. Attenuation measures the energy lost as waves propagate through the geology.

Further, the values of Vp and Vs throughout the volume can be used to calculate Poisson’s ratio, and Bulk, Young’s, Lambda and shear moduli throughout the volume. These proprieties are utilized in the context of rock physics relationships to identify the effects of fluids, fractures, porosity, and permeability on seismic velocity [46]. Following [46], several typical interpretations of porosity, permeability and saturation can be made from observable microearthquake data. Comparisons are made relative to normal geology at similar depths and temperatures, meaning geology that has a monotonic increase in velocity and Q as a function of depth, and saturation, porosity and temperature that is considered average for the geologic condition of the study area [46].

Figure 11 shows a cross-section of a likely reservoir in southern Taiwan [43]. Black dots are micro-earthquakes. Low values of shear modulus (red) indicate a soft geology [43], likely due to fractures and may indicate the location of the reservoir. Almost no earthquakes occur in the soft geology. This shows how faulting can create permeability. The faults create pathways for fluids, and highly fracture the geology. Here, the conjugate faulting matches the flanks of anticlines caused by the orogeny deformation [43].

Figure 11.

Shear modulus. Low values likely indicate fractures. From [43].

Figure 12 shows Qs for the same volume [45]. The low values of Qs can indicate the existence of water. Shear wave propagation is not affected by liquid, so this figure is independent of the shear modulus. Together, the fractures and saturation are a good indication of a reservoir. Blue lines indication the location of possible faults, which may provide pathways for water.

Figure 12.

Qs. Low values (orange) likely indicate saturation. From [45].

7. Induced seismicity

During EGS geothermal reservoir creation or organic stimulation, rocks may slip along pre-existing fractures and produce microseismic events. Researchers have found these microseismic events, also known as induced seismicity, to be a very useful diagnostic tool for accurately pinpointing where fractures are re-opened or created, and characterizing the extent of a reservoir. In almost all cases, these events occur in deep reservoirs and are of such low magnitude that they are not felt at the surface. Although induced seismicity data allows better subsurface characterization, GTO also understands public concern. With this in mind, US DOE led an effort to create a protocol for addressing induced seismicity associated with geothermal development, which all US DOE-funded EGS projects are required to follow [50]. This work was informed by panels of international experts and culminated in an International Energy Agency- accepted protocol in 2008. The protocol was updated in early 2011 to reflect the latest research and lessons learned from the geothermal community. In addition, in June 2012, the US National Academy of Sciences (NAS) issued Induced Seismicity Protocol in Energy Technologies [51]. The report found that geothermal development, in general, has a low potential for hazard from induced seismicity. The NAS report cited the DOE Induced Seismicity Protocol as a best practice model for other subsurface energy technologies.

Figure 13 is a demonstration of induced seismicity due to cold water released by gravity flow over a month period of time in four wells at the Geysers, California geothermal reservoir [52]. The white dots are the earthquakes, including the natural and human-induced earthquakes. Vs tomography results are shown as backdrop. In Figure 13a, the well at far left, water release was increased just prior to the month. For the well second from left, water was released only for two days at the end of the month. Seven earthquakes occurred after water release was started and no earthquakes occurred prior. For the well third from left water was released at a low rate during the month. The seismicity was relatively little near the bottom of the well. The well at the far right is not in hot geology, and no earthquakes are observed.

Figure 13.

Demonstration of induced seismicity due to the water release in hot geology.

Figure 13b shows the effect of water injection at four wells for the month immediately following. The well on the far left has ceased generating earthquakes. The well second from left has created many induced earthquakes due to the ongoing release of water. The third well from the left has increased water flow and shows an increase in seismicity at the bottom of the well. The well at the far right continues to show no earthquakes.

The case study presented above illustrates that by continuously monitoring seismic activity and modifying water injection accordingly, the occurrence of induced seismicity can be mitigated.

8. Discussion and conclusions

Power produced from geothermal energy is an important component of a renewable energy demand because it is continuously available, as opposed to wind and solar, which are intermittent in availability.

The recent tectonic evolution of Taiwan created ideal conditions for geothermal resources; heat, water and permeability. The island formed as a result of the transition from subduction of the Eurasian Plate under the Philippine Sea plate to active collision. The Central Range has been the most dramatic manifestation of this collision. It has experienced the highest mountain uplift in the world, as much as ~26 mm/year. This active orogeny creates unusually high geothermal gradients by exhumation of the warmer material from depth and by strain heating. As a result, temperatures reach up to ~200 degree C. Volcanoes in the northern tip of Taiwan provide an additional source of heat. Favorable fluid flow from meteoric waters and permeability from seismicity and faulting results in exploitable geothermal systems near the surface. These systems can potentially provide geothermal power generation throughout the whole Island.

The average land heat flow in the world is about 30 mW/M². If there is a local heat flow value greater than the average, then there is a good potential for geothermal development if permeability and fluids also exist. Song and Liu [37] identified 11 major geothermal resource regions based on anomalous geothermal gradients >35°C/km (Figure 6). These 11 high anomalous areas of geothermal gradient are the potential areas of geothermal development. Hot springs are prevalent throughout the 11 areas (Figure 3), indicating permeability and fluids. These resource areas occur either along the central range with the high rate of seismicity that sustained the 1999 M = 7.8 Chi-Chi earthquake (II, III, VII, Figure 6), along the Philippine fault within the Longitudinal Valley (IV, V, VI); both of which provide ample permeability for hot fluid circulation. Area I is within the Tatun volcanic area and permeability and fluid flow has been observed from exploratory wells [41, 53]. Area VII has high heat flow and hot springs, but no further confirmation of permeability has been identified. Taiwan has about 2500 mm of rain each year [38], sufficient for replenishing reservoirs.

In Taiwan, orogenic reservoirs are typically located at a shallow depth of 2–3 kilometers in the Central range and volcanic zone where the geothermal gradient is greater than 30^o C/km. The development depth of enhanced geothermal away from the Central range is usually greater than 3000 m where the geothermal gradient is approximately 30 mW/M². Exploration is necessary to determine whether there is an organic reservoir or a candidate for an engineered system. Taiwan is currently conducting small-scale exploration to potentially exploit potential geothermal resources [39].