Sustainable Yield and Its Assessment of Geothermal Reservoirs in China

Transcription

1 GRC Transactions, Vol. 37, 213 Sustainable Yield and Its Assessment of Geothermal Reservoirs in China Fengxin Kang Shandong Provincial Bureau of Geology and Mineral Resources, Jinan, China Keywords Geothermal water; water level decline; temperature drop; allowable drawdown; sustainable yield; sustainable development Abstract Aiming sustainable development of geothermal resources, this paper explores the concept of sustainable yield of geothermal wells and reservoirs, i.e. geothermal exploitation over a specified timeframe without causing: decrease water level (pressure) below the designed, impact the normal exploitation of existed geothermal extraction wells, water quality and temperature lower than the allowable limits, induce harmful environmental issues. The prerequisite of calculating sustainable yield is to determine a rational maximum allowable drawdown of the production wells in the geothermal reservoir, within a time frame. For a new discovered reservoir, which has not been heavily extracted, it is essential to build a conceptual and numerical model to carry out sustainable yield evaluation based on interference pumping test and water level response observation. Whereas for a developed geothermal reservoir, it is indispensable to evaluate sustainable yield in view of long-term water level response and the constraints of unacceptable environmental geothermal issues. Four case studies have been presented in this paper, one is for a heavily developed sedimentary geothermal reservoir with continuously water level declining, the second is for a newly discovered sedimentary geothermal reservoir, the third and fourth are for an overexploited fractured geothermal reservoir with temperature and TDS drop. The corresponding distributed and lumped parameter models have been successfully built to evaluate sustainable yield. Geothermal Development in China China is located in the eastern part of Eurasia continent. The coastal area in east China, which borders on the Pacific Ring of Fire Fig. 1), is a zone of frequent volcanic eruptions, earthquakes and geothermal activities, so there are a lot of hot springs in Liaoning, Shandong, Fujian, Guangdong, Taiwan and Hainan Provinces. In addition, because China s southwest border lies on the eastern part of Mediterranean-Himalayan belt, hot springs are common in Tibet and western Yunnan, and there are many hightemperature geothermal manifestations, such as boiling springs, boiling fountains and geysers. More than 3, hot springs lie mostly in these two belts. Figure 1. Pacific ring of fire: shows the techtonically active shores of the Pacific Ocean, where many of the earth s vocanoes, earthquakes and geothermal activities occur, associated with the earth s palte boundaries(red points: volcanoes; yellow points: earthquakes, black lins:plate boundaries)(from NIWA, 213). Meanwhile, there are more than 3,5 geothermal wells drilled in China, most of which lie in the large and medium-sized basins in northern and the eastern plains region. High temperature geothermal resources in China are distributed in the zone from Tibet to West Yunnan, which have been used for power generation in Yangbajain, Tibet, with an operating installed capacity of MWe. Geothermal power satisfied 5% electricity demand in Lhasa, the capital of Tibet, even 6% in winter (Fig. 2, Fig.3). China is rich in, especially low-medium temperature geothermal resources. There are more than 3 natural geothermal springs and 35 geothermal wells distributed in 843

2 The total annual direct use of geothermal energy reached 2,932 GWh(75,348TJ/yr) in 29(Lund et al., 21), indicating a 66% increase over 25(Lund et al., 25), which ranked first in the world. The geothermal water production reached 5 million m 3. The space heating area in China has reached 35 million m 2 by using geothermal water, and 1.7 million m 2 by using geothermal heat pumps (Zheng et al., 21). Environmental Geothermal Issues Due to the heavy exploitation of geothermal water and irrational layout of exploitation, overdevelopment of geothermal resources has occurred in China, and induced some corresponding environmental geothermal issues, including water level (pressure) declining continuously, cold water intrusion and temperature and TDS drop, waste water pollution. Figure 2. Locations of the study areas in China. Continuous Geothermal Water Level (Pressure) Declining Corresponding to the heavy development of geothermal water in recent years, the water level (pressure) in the reservoir has been decreasing dramatically such as in reservoirs of Beijing, Tianjin, Xi an and Xianyang. Beijing Urban Geothermal Field The geothermal fields beneath the Beijing Plains area are associated with unique Mesoproterozoic basement aquifers of high permeability (Fig. 4) (Hochstein et al., 1988, 22). The main aquifer appears to be coherent beneath the whole Beijing Graben where it can be found typically between 5 and 4, m depth over an area of the order of 2,76 km 2, with temperatures ranging from 4 to 9 C. Figure 3. Geothermal power generation in Yangbajain. the country. Low-medium temperature geothermal resources are directly used in various aspects in China: space heating, bathing, balneology, fish farming, sanatorium, swimming pool and greenhouse, etc. Geothermal energy is one of the energy resources that can be used in a sustainable manner, as well as to mitigate climate change(axelsson, 21). With geothermal energy becoming increasingly more competitive with fossil fuels and the environmental benefits associated with renewable energy resources better understood, development of this natural heat from the earth has been accelerated in China for the last 2 years. Figure 4. Satellite image of Beijing City showing the plains area where geothermal fields located (from Liu, 28). 844

3 2 Kang Shortly after the discovery of the Beijing urban geothermal field, thermal water was already abstracted and utilized with the completion of the first geothermal well in Since then, the number and depth of geothermal wells have increased gradually between 1972 and 198 thermal water was mainly used for industrial processing and minor space heating. Since 198 utilization shifted more towards major space heating and bathing which resulted in pronounced seasonal abstraction peaks. As shown in Fig. 5, in the southeast urban geothermal field annual production of thermal water had already reached 3.5 Mm 3 /yr in 1982, attaining a peak of 5 Mm 3 /yr in 1985 and declining gradually to 3.5 Mm 3 /yr at present (Hochstein et al., 22). Continuous production of thermal water led to a decline in aquifer pressure resulting in a drop of the water level. The changes of the water levels in Wumishan dolomite aquifer, as monitored in well JR-7 in the center of Beijing urban field, are also shown in Fig. 5. The biggest water level drawdown has reached 7 m, with annual water level decrease of m. Monthly production rate (X1 3 m 3 ) Monthly production rate Water level Figure 5. Water level variations versus monthly production in SE urban geothermal field of Beijing (adapted from Liu, 28). After the discovery of Up to 28, 4 geothermal wells have been drilled in the ten geothermal fields of Beijing. Among them, 178 are presently producing wells, 18 are reinjection wells, 9 are monitoring wells, and 195 are to be used. The total production rates in 28 wre around 8 Mm 3. Tianjin Geothermal Field The Tianjin geothermal field is a typical low-temperature system, which is located in the middle-lower reaches of the Haihe River on the North China plain (Fig. 2). It covers an area of 87 km 2, approximately 77% of the total Tianjin area. Geothermal water is mainly accumulated in lower Paleozoic Cambrian- Ordovician and medium Proterozoic Jixiannian karst aquifers, as well as Tertiary sandstone aquifers, with the temperatures ranging from 4 to 13 C. In early 1936, the first geothermal well was drilled in the center of Tianjin urban area. By the end of 27, there were 256 production wells and 38 reinjection wells in Tianjin, with total production rates of 24.5 Mm 3, and reinjection rate of 4.62 Mm 3. In 27, the space heating areas have covered 12 Mm 2, accounting for 8% of the total heating area in winter. Fig. 6 illustrates the history curve of water level drawdown and geothermal water production of the dolomite and limestone reservoir in Tianjin urban area from 1992 to 22. Since 1997, the annual water level drawdown has been over 3 m, and even got to 1 m in 22. At present, the depth to the water level varies between 4 m and 12 m in different areas. The area of water 4-4 Water level (m) Annual production rate (X1 6 m 3 ) Figure 6. Relation between water level drawdown and geothermal water production of the dolomite reservoir in Tianjin urban area. Area of water level deeper than 9 m (km ) Annual production rate 45 Annual water level drawdown Year Figure 7. The area of water level deeper than 9 m in Tianjin. level deeper than 9 m has reached 554 km 2 in 27 (Fig. 7). Now, the geothermal water level is still decreasing at a rate of 6-9 m annually. Temperature and TDS Drop Corresponding to the water level drawdown, obvious temperature and TDS decrease occurred in some geothermal reservoirs, especially in natural warm spring areas due to shallow colder groundwater or surface water inflow. Q,TDS T( C) TDS(x1mg/L) Q(m 3 /d)) Figure 8. Correlation among production (Q), temperature (T) and water quality (TDS) of Tangtou geothermal spring, Shandong, China Auunal water level drawdown (m) T

4 Tangtou, one of the 18 natural geothermal springs in Shandong province of Eastern China (Fig. 2), its temperature and TDS have decreased since 197 s due to large scale development (Fig. 8). Before 196 s, the free flow rate was 388 m 3 /d, with temperatures ranging around 7 C and TDS around 4.5 g/l. In 2, the total production from geothermal wells reached 955 m 3 /d, its temperature dropped to 48 C, and TDS to 1.9 g/l. Waste Water Pollution Generally speaking, geothermal is renewable and clean energy, but the geothermal waste waters have higher salinity and contain pollutants as Fluorine (F), Boron (B), Mercury (Hg), Arsenic (As). Parts of the pollutant contents exceed the national disposal water quality standards for drinking, irrigation and aquaculture water. Meanwhile, cascaded and comprehensive use of geothermal water in China is not so common. As a result, the discharged water usually has higher temperature with higher thermal energy. Sustainable Yield of Geothermal Reservoir The prerequisite for sustainable geothermal extraction is to determine the sustainable yields of the production wells and geothermal reservoir. The sustainable yield can be defined as geothermal exploitation over a specified timeframe without causing: (1) decrease water level (pressure) below the designed, (2) impact the normal exploitation of existed geothermal extraction wells, (3) water quality and temperature lower than the allowable limits, (4) induce harmful environmental issues. Based on the behaviors of geothermal development, particularly focused on the environmental geothermal issues, the author believes that the most fundamental indicator for sustainable yield of geothermal water is maximum allowable water level (pressure) drawdown. The reason for this is that most of the environmental geothermal issues are induced by too deep drawdown of water level (pressure). To be brief, the sustainable yield of geothermal water is the maximum allowable yield when the actual drawdown doesn t exceed the maximum allowable drawdown within a time frame. Accordingly, the prerequisite to determine the sustainable yield is to determine the maximum allowable drawdown. For a newly discovered geothermal reservoir, which has not been heavily extracted, it is essential to build a conceptual and numerical model to carry out sustainable yield evaluation based on interference pumping test and water level response observation, as well as response predicting for the long-term behaviors of wells and reservoirs, including water level changes and its side effects. Detailed numerical modeling is an effective method for production potential calculation and evaluation by calculating and predicting water level behaviors for different future production scenarios. It can take interference of different wells into account, also consider the reservoir boundary conditions-open or close or in between, which has drastically different long-term behaviors individually. To fulfill this aim, the maximum allowable water level drawdown should be determined carefully as the precondition of the sustainable yield evaluation. In other words, sustainable yield is drawdown dependent. Whereas for a developed geothermal reservoir, it is essential to evaluate sustainable yield in view of long-term water level (pressure) response over production and the constraints of unacceptable environmental geothermal issues. So, the importance of long-term measurements of geothermal water level (pressure) should be highlighted. It is believed that the sustainable development of geothermal resources can be achieved if the actual production is based on the sustainable yield, i.e. the water level, temperature and water quality will be maintained in the allowable limits at the end of time frame. Case Study 1: Newly Discovered Sedimentary Geothermal Reservoirs Sedimentary geothermal reservoirs, which are widely distributed in the plain and basin areas of China, are characterized by large areal extent, rather homogeneous aquifers, quite productive wells and great energy potential (Kang, 2). Considering constraints in the sedimentary geothermal reservoir, including (1) setting depths of submersible pumps, (2) design of the production wells, (3) risk of colder water inflow, (4) subsidence, (5) the maximum allowable drawdown is determined as 1-3 m within 1 years. The Dezhou geothermal reservoir is located in the alluvial plain dominated by the Yellow River, which is within the North China Sedimentary Basin (Fig. 2). Dezhou, a city situated in the northwest part of Shandong Province, has a population of 3, and lies approximately in the centre of the geothermal area. The Dezhou geothermal reservoir is a low-temperature sedimentary reservoir yielding water with a temperature between 46 and 58 C. Two successful productive wells have been drilled into the reservoir since The emphasis on geothermal development has been in the area of direct-utilization, such as for space heating, swimming pools and balneology. Owing to its very low actual production, 2.9 l/s in 1999, the Dezhou geothermal reservoir, on the whole, was still in its natural state. However, cold winters coupled with growing concerns over greenhouse gas emissions suggest that the geothermal waters will be developed intensively in the future. For the purpose of rational exploitation of the geothermal resource, it is essential to carry out a reservoir evaluation and estimate the sustainable yield, by predicting water level response in the reservoir. Geological Background The Dezhou geothermal reservoir is situated within the Dezhou depression. It is bounded by the Bianlinzhen fault on the east, the Cangdong fault on the west, the Xiaoyuzhuang fault on the south, and the Xisongmen fault on the north. All of these faults appear to act as permeable boundaries, which are presented in Fig. 9. Some other faults, such as the Jianhe fault, intersect the Dezhou reservoir, and then result in anisotropic permeabilities of the reservoir. 846

5 1 2 3 K+ J 4 C a ng dong Uplift EW DR1 C a ngdong Fa ult DR2 Depth(m) Q Nm 1 Ng Es 2 Ek Jia nhe Fa ult Bia nlinzhen Fa ult K+ J 3 K+ J C he ng 4 Depre ssion ning Uplift Figure 9. Tectonic cross-section of the Dezhou depression. The Dezhou geothermal reservoir is a sedimentary reservoir with heat-flow dominated by conduction. It is believed that the reservoir exists because of the occurrence of highly permeable sedimentary layers at great depth, and an above average geothermal gradient; as well as because of the faults and fractures. The cap rock is upper Minghuazhen formation of Neogene age. The upper Minghuazhen formation, with a thickness of 9 m, is composed of argillite and sandy argillite with interbedded sandstone. The Dezhou geothermal reservoir is located within the Guantao formation of Neogene age, with a depth ranging from 135 to 165 m and a thickness of 3-48 m. The main production aquifer of the reservoir is comprised of sandstone and conglomerate, covers an area of 169 km 2 and has a thickness of m. The production aquifer has high porosity (24-3%) based on cores. In the natural state, the wells are artesian, with an artesian pressure of 7-8 m and a free-flow rate of l/s. Production rates of single wells are l/s at pump depths of about 5 m, and temperatures C at depths between 135 and 16 m. Conceptual Model Conceptual model is the fundamental element of reservoir modelling. The conceptual model of the Dezhou geothermal reservoir may be briefly delineated as follows: (1) Reservoir type: low-temperature sedimentary sandstone reservoir, conduction-dominated; (2) Boundary: permeable faults boundaries; (3) Production aquifer: confined Neogene Guantao Formation, with a thickness from 16 to 18 m at a depth between 135 and 165m, and covering an area of 169 km 2 ; (4) Cap rock: upper Minghuazhen formation of Neogene period, composed of argillite and sandy mudstone; (5) Underlying rock: Eogene Dongying Formation, composed of argillite, fine sandstone, and siltstone; (6) Recharge: meteoric origin. Distributed Parameter Modelling On the basis of its conceptual model, the Dezhou geothermal reservoir can be outlined as horizontal and homogeneous, with a constant thickness and an infinite areal extent. This kind of reservoir is in good agreement with the prerequisites of a simple analytical distributed parameter model. In the distributed parameter model, locations, productions, injections, and observations of each individual wells are taken into account. In other words, the behaviour of individual wells and the interferences among different wells can be simulated by the distributed model. The distributed parameter computer code VARFLOW (EG&G Idaho Inc. et al., 1982) is based on the Theis model as follows: where Δp(t) = Pressure change at time t due to the flow rate q(τ ), τ n < t < τ n+1 (bar); μ = Dynamic viscosity of the fluid (Pa.s, kg/m/s); k = Permeability (Darcy, 1-12 m 2 ); h = Reservoir thickness (m); τ n = Time at which the flow starts (s); τ n+1 = Time at which the flow stops (s); q(τ) = Volumetric flow rate at time τ (l/s); r = Distance between the observation well and the production/injection well (m); η = The hydraulic diffusivity ( k / µc t ) (m 2 /s); c t = c w ϕ + c r (1 ϕ), total compressibility of water saturated formation (1/Pa); c w = Compressibility of water (1/Pa); c r = Compressibility of rock matrix (1/Pa); ϕ = Reservoir porosity. VARFLOW calculates pressure changes in response to water production/injection from/into an idealized reservoir system. Based on the actual conditions of the Dezhou reservoir, anisotropic permeabilities are chosen, and the initial state of reservoir pressures prior to productions is assumed as constant. The parameters (transmissivity and storage coefficient) are varied until a satisfactory match was obtained. Fig. 1 represents the match between the observed and simulated pressure in well DR1 by using the data of short-term well test during Mar. 28 Apr. 4, It is obvious that the match is quite good. The parameters of the reservoir are estimated as follows: X-direction transmissivity: Tx = k x h / µ = m 3 / Pa.s ; Y-direction transmissivity: T y = k y h / µ = m 3 / Pa.s ; Storage coefficient: S = c t h = m / Pa. A moderate anisotropy, k y / k x = 5, is incorporated in the model, which means that the permeability in the north-south direction is 5 times that in the east-west direction. The north-south direction coincides with the concentrated runoff zone direction of the groundwater in the Dezhou geothermal reservoir. From the above results, the average permeability-thickness and permeability of the reservoir are calculated as (h=17 m): Average kh = k x h = 5 Darcy-m k y h = 25 Darcy-m (k x h)(k y h) = 1118 Darcy-m k x = 2.94 Darcy k y = 14.7 Darcy Average k = k x k y = 6.6 Darcy 847

6 Pressure(bar) Production(l/s) Measured pressure Simulated presssure 3/28/97 3/31/97 4/2/97 4/5/97 Time(mm/dd/yy) Figure 1. Comparison between observed and simulated (Theis-model, VARFLOW) pressure changes combined with production during testing of well DR1 in In view of the fact that the production response of a reservoir is chiefly manifested as water level drawdown, the models established here may be used to evaluate the sustainable yield of the reservoir by calculating and predicting water level performance for different future production scenarios. The prerequisite of calculating production potential is to determine a rational maximum allowable drawdown of the production wells in the reservoir, since it determines the economic production potential directly. In other words, the sustainable yield is restrained by the maximum allowable drawdown of the reservoir. Sustainable Yield of the Dezhou Geothermal Reservoir Considering constraints in the Dezhou geothermal reservoir, including the setting depths of well pumps, design of the production wells, risk of colder water inflow, and especially the land subsidence in the area, the maximum allowable drawdown is defined as 1 m. On the basis of the maximum allowable drawdown and the established distributed parameter model, the water level predictions were calculated for two different production scenarios: (1) Scenario I: the production during pumping test for wells DR1 and DR2 is maintained for the next ten years, i.e.: Annual average production: 62 l/s; Production in heating season (Nov. Mar.): 12 l/s; Production in non-heating season: 1 l/s. The water level predictions, calculated by the distributed parameter model, are presented in Fig. 11, which shows the calculated water level changes in well DR1 for the next ten years. (2) Scenario II: besides the wells DR1 and DR2, four additional production wells are included according to geological conditions and the requirements of the Dezhou municipality. The production is increased to: Annual average production: 22 l/s; Production in heating season: 36 l/s; Production in non-heating season: 12 l/s. The water level predictions, calculated by the distributed parameter model, are also presented in Fig. 11, which shows the water level changes in well DR1 for the next ten years. The well Water level(m) 4 2 Scenario I Scenario II /31/1 7/31/13 11/1/16 3/31/2 Time(mm/dd/yy) Figure 11. Predictions of water level trends of the next 1 years in well DR1 for production scenario I and II for the Dezhou geothermal field, calculated by VARFLOW. DR1 is still located in the centre of the depression cone of the water level (Fig. 12). Fig. 12 illustrates the water level contours at the end of the nest 1 years, calculated by the distributed model, combined with a three-dimensional presentation of the water level surface. Fig show that the greatest anticipated water level drawdown for scenario I is 58 m, whereas for scenario II it is 99 m. In light of these results, and considering the maximum allowable drawdown of 1 m, the sustainable yield of the Dezhou geothermal reservoir is evaluated to be 22 l/s on the average for the next ten years, or 6.9 Mm 3 per year. The allowable maximum production in heating seasons is 36 l/s. Case Study 2: Heavily Developed Sedimentary Geothermal Reservoir Water level (m) Figure 12. Predicted water level contours (m) (top) and 3D water level surface (bottom) in Dezhou at the end of the next 1 years for production scenario II, calculated by VARFLOW. Long term abstraction of geothermal water has reduced the pressure in the dolomite and limestone aquifers over the

7 whole Plain area which is indicated by the decline of the water (piezometric) levels in monitoring wells. The urban geothermal field is one of the ten in Beijing Plain area, has been developing since 1971, and continuous water level (pressure) and production monitoring data are available since then. The water level decline has up to 7 m, equivalent to 7. bar pressure drop. Lumped parameter models have been extensively used to simulate data on water level and pressure changes in geothermal systems in Iceland (Axelsson, 2). An automatic non-linear iterative least-squares technique for estimating model parameters, which tackles the simulation as an inverse problem, was applied in the simulation. Being automatic it requires very little time compared to other forward modelling approaches, in particular detailed numerical modelling. Lumped parameter models can simply be considered as distributed parameter models with a very coarse spatial discretization (Bodvarsson et al., 1986). The method presented here tackles the modelling as an inverse problem that requires far less time than direct, or forward modelling, where the interactions are done manually. This makes the lumped parameter simulations highly cost effective. To tackle the simulation as an automatic inverse problem, a powerful and effective computer code LUMPFIT has been developed. A general lumped model is shown in Fig. 13, which consists of a few tanks and flow resistors. The water level or pressure in the tanks simulates the water level or pressure in different parts of a geothermal system. The pressure response (p) of a general open lumped model with N tanks, to a constant production (Q) since time t=, is given by the equation: geothermal system. When using this method of lumped parameter modelling, the data fitted (simulated) are the water level data for an observation well inside the well field, while the input for the model is the production history of the geothermal field in question. A first guess of the lumped model parameters is made and then, consequently, the parameters are changed by the automatic iterative process described above until a satisfactory match (in the least squares sense) for the selected model is obtained. There are no a-priori assumptions made on the nature and geometry of the reservoir (Axelsson, 2). Figure 13. Schematic diagram of lumped parameter model (Axelsson, 2). The pressure response of an equivalent N tank closed model is given by the equation: The coefficients A j, L j and B are functions of the storage coefficients of the tanks (κ j ) and the conductance coefficients of resistors (σ j ) in the model. The resistors, controlled by permeability of rocks, simulate the flow resistance in a reservoir. The first tank simulates the innermost part of a geothermal reservoir, i.e., it represents the active well field; while the second and third tanks simulate outer and deeper parts of a system, i.e., they act as recharge parts from either deeper or outside parts of the main reservoir. If the third tank is connected by a resistor to a constant pressure source, which supplies recharge to a geothermal system, the model is therefore open. Otherwise, without the connection to a constant pressure source the model would be closed. An open model acts as optimistic, since equilibrium between production and recharge is eventually reached during long-term production, causing water level drawdown to stabilize. In contrast, a closed model may be considered as pessimistic, since no recharge is allowed for such a model and the water level declines continuously as production proceeds. Hot water is pumped out from the first tank, which causes the pressure or water level in the model to decline. This in turn simulates the decline of pressure or water level in the real Figure 14. The production history and water level response of the southeastern Urban geothermal field in Beijing with the water level history simulated by a lumped parameter model (squares = measured data, line = simulated data) (from Axelsson et al., 25) Figure 15. Predicted water level variations in the southeastern Urban geothermal field of Beijing over the scenario of annul production = 6 Mm 3 /yr. The simulated and monitored water level changes of southeastern urban geothermal field in Beijing over the period from 198 to 22 combine with the production rates are shown in Fig. 849

8 14 by using a lumped parameter model of Liu et al. (22). Fig. 14 presents a very good agreement between observed and fitted data. Consequently, this lumped parameter model can be used to predict water level responses over different production scenarios for a 1-year time frame. As shown in Fig. 15, the sustainable yield of the southern urban geothermal field in Beijing is evaluated to be 6 million m 3 annually, with the maximum allowable water level drawdown of 1 m for the next 1-year. Case Study 3: Fractured Geothermal Reservoir Fractured geothermal reservoirs in China are mainly distributed in the eastern and southeastern coastal zones of China, which are controlled by neotectonisms and large- and deep faults, with convective heat flow. Their main aquifers are granite and metamorphic rocks, and the natural hot springs are the major geothermal manifestations, with temperatures varying from 4 to 12 C. Considering main restrictions in the fractured geothermal reservoir, including (1) Risk of colder water inflow (surface water, ground water, and sea water), (2) Temperature and TDS decrease, the maximum allowable drawdown is defined as 5-5 m, based on the critical water level when harmful water temperature and TDS drop appear for different geothermal reservoirs. For Tangtou fractured geothermal reservoir, the maximum allowable drawdown is defined as 11 m, which means that when the water level is deeper than 11 m, obvious water temperature and TDS drop will occur in the reservoir-temperature below 6 C and TDS below 2.5 g/l (Fig. 8). Corresponding to the maximum allowable water level drawdown, the sustainable yield is determined as 5 m 3 /d. That is to say, when the production in the Tangtou geothermal reservoir is less or equal 5 m 3 /d, the water temperature will be higher or equal 6 C and TDS will be higher or equal 2.5 g/l. Yujiatang fractured geothermal reservoir is located in Shandong Peninsula. Based on the long-term observation data, the relationship between water levels and temperatures has been obviously shown in Fig The fitted equation between Temperature (T) and Water level depth (D) is evaluated as: T = exp( D) (Coef. of determination, R-squared =.84) Figure 16. Effects of water level changes on temperature in Yujiatang geothermal reservoir, Shandong Peninsula. When the water level is deeper than 12 m, the water temperature is lower than 38 C, which decrease the utilization Temperature( o C) Water level depth (m) Figure 17. Correlation of temperatures with water levels in Yujiatang geothermal reservoir, Shandong Peninsula. value remarkably. Accordingly, the maximum allowable drawdown is determined to be 12 m, and the corresponding sustainable yield is evaluated to be 633 m 3 /d. Sustainable Yield With Reinjection of Geothermal Reservoir From the point of view of recharge ability of geothermal water, the geothermal systems can be classified as either closed systems with limited or no recharge, or open systems where recharge equilibrates with mass extraction. It is believed that most sedimentary geothermal reservoirs are of the feature of closed system with continuously water level (pressure) declining in response to long-term large-scale development; whereas most of the fractured geothermal reservoirs are of the feature of open system with dynamic stabilization of water level (pressure) if the production rates are based on the sustainable yield. In response to no or very limited recharge of closed sedimentary geothermal reservoirs, water level (pressure) will continuously declining corresponding to the mass extraction until to the exhaustion of the reservoir. For the purpose of counteracting water level (pressure) declining and then stabilizing the production capacity of the geothermal field through the maintenance of the water level (pressure), reinjection is necessary for most of the sedimentary geothermal reservoirs. Meanwhile, reinjection has such benefits as: reducing waste water disposal from power plant and return water from direct use of geothermal water for environmental reasons - thermal and chemical pollution, to enhance energy extracting from aquifer rock matrix along flow paths from reinjection wells where most (>9%) of the potentially useable geothermal energy is stored, to offset against subsidence caused by production induced water level (pressure) decline, as well as to enhance or revitalize surface thermal manifestations such as hot springs and fumaroles (Bromley et al., 26). Reinjection is believed to have started as soon as in the late 196 s, both in high and low-temperature geothermal fields. The first known instance of reinjection into a high-temperature geothermal system was in the Ahuachapan field in El Salvador, starting in 1969 (Stefánsson, 1997). Low-temperature reinjection also started in the Paris Basin in 1969 and has continued ever since. During the 197 s the number of reinjection operations started picking up and reinjection experience really started growing (Axelsson, 28). 85

9 In Beijing, the earliest geothermal reinjection experiments were started in the urban geothermal field in 1974 and In the heating period, there were 6 reinjection wells for injecting waste water from 8 production wells in the Xiaotangshan Geothermal field, which is one of the 1 geothermal fields in Beijing Plain area, about 3 km north of Beijing centre. The total quantity of reinjection was 1.32 million m 3, accounting for 56.6 % of the annual production in the field. As a result, the water level in the geothermal field not only stopped declining, but also started to rise since 25 (Fig. 18). In the winter of 25-26, the water level was about 5 m higher than that in (Liu et al., 28). This shows clearly that reinjection is the most effective countermeasure for ensuring sustainable use of geothermal energy in Beijing, as well as other geothermal fields in China. Consequently, the sustainable yield will be largely increased for most of the geothermal fields. Water level (m) Conclusions Month Figure 18. Water level variations in Xiaotangshan geothermal field of Beijing during reinjection from 23 to 26. The prerequisite for sustainable geothermal extraction is to determine the sustainable yield of production wells in the geothermal reservoirs. The sustainable yield is defined as geothermal exploitation over a specified timeframe without causing such issues as: decrease water level (pressure) below the designed, impact the normal exploitation of existed geothermal extraction wells, water quality and temperature lower than the allowable limits, induce harmful environmental issues. The most fundamental indicator for sustainable yield of geothermal water is maximum allowable water level (pressure) drawdown, and the sustainable yield is the maximum allowable yield when the actual drawdown doesn t exceed the maximum allowable drawdown within a time frame. For a new discovered reservoir, which has not been heavily extracted, it is essential to build a conceptual and numerical model to carry out sustainable yield evaluation based on interference pumping test and water level response observation, as well as response predicting for the long-term behaviors of wells in the reservoir, including water level changes and its side effects. For a developed geothermal reservoir, it is essential to evaluate sustainable yield in view of long-term water level (pressure) response and the constraints of unacceptable environmental geothermal issues. So, the importance of long-term measurements of geothermal water level (pressure) should be highlighted. If the actual production is based on the sustainable yield, it is believed that the sustainable development of geothermal energy can be achieved, i.e. the water level, temperature and water quality will be maintained in the allowable limit at the end of the time frame. Reinjection is the most effective countermeasure for ensuring sustainable use of geothermal energy. References Axelsson G (2) Sedimentary geothermal systems in China and Europe. In: Long-term monitoring of high- and low-enthalpy fields under exploitation. WGC2 short courses, Kokonoe, Japan, Axelsson G, Björnsson G and Quijano J (25) Reliability of Lumped Parameter Modeling of pressure changes in geothermal reservoirs. Proceedings of the World Geothermal Congress 25, Antalya, Turkey, April, 8 pp. Axelsson G, Stefánsson V, Björnsson G (25) Sustainable utilization of geothermal resources for 1 3 years. Proceedings World Geothermal Congress 25, Antalya, Turkey, April, 8 pp. Axelsson G (28) Importance of Geothermal Reinjection. Proceedings, workshop for Decision Makers on Direct Heating Use of Geothermal Resources in Asia, Tianjin, China, 28, Axelsson G (21) Sustainable geothermal utilization-case histories; definitions; research issues and modeling, Geothermics xxx (21) xxx -xxx, Elsevier, London (Article in press). Bromley, C.J., M. Mongillo and L. Rybach, 26: Sustainable utilization strategies and promotion of beneficial environmental effects Having your cake and eating it too. Proceedings of the New Zealand Geothermal Workshop 26, Auckland, New Zealand, November, 5 pp. EG&G Idaho Inc., Lawrence Berkeley Laboratory (1982) Low- to moderate-temperature hydrothermal reservoir engineering handbook. Idaho Operations Office, USA, report IDO-199, Appendix E, 4 pp. Hochstein MP, Yang Zhongke (1988) The Beijing geothermal system, P.R. China: natural state and exploration modeling study of a low temperature basement aquifer system. Proceedings 13th workshop on reservoir engineering, Stanford University, Hochstein MP, Zheng KY (22) Assessment of geothermal resources in Beijing plains area. Proceedings of the International Symposium on Geothermal and the 28 Olympics in Beijing, Beijing, October 22, Kang FX (2) Assessment of sedimentary geothermal resources in Dezhou, China and Glalanta, Slovakia. Reports of the United Nations University geothermal training programme in 2, Liu JR, Pan XP, Yang YJ, Liu ZG, Wang XL, Zhang LH, Xu W (22) Potential assessment of the Urban geothermal field, Beijing, China. Proceedings of the International Symposium on Geothermal and the 28 Olympics in Beijing, Beijing, October 22, Liu JR, Wang K (28) Experience of Geothermal Reinjection In Beijing And Tianjin. Proceedings, workshop for Decision Makers on Direct Heating Use of Geothermal Resources in Asia, Tianjin, China, 28, Lund JW, Freeston DH, Boyd TL (25) Direct application of geothermal energy 25: worldwide review, Geothermics 34, Elsevier, London, UK, Lund JW, Freeston DH, Boyd TL (21) Direct utilization of geothermal energy 21 worldwide review. Proceedings of the World Geothermal Congress 21, Bali, Indonesai, April. Manfred P, Hochstein and Keyan Zheng, 22: Assessment of Geothermal Resources in Beijing Plains area. Proceedings of the International 851

10 Symposium on Geothermal and the 28 Olympics in Beijing, Beijing, October 22, NIWA-National Institute of Water and Atmospheric Research and National Geophysical Data Center, National Oceanic and Atmospheric Administration (213) Pacific Ring of Fire. en/interactive/5581/pacific-ring-of-fire Stefansson V (1997) Geothermal reinjection experience. Geothermics 26, Elsevier, London, UK, Zheng KY, Han ZS, Zhang ZG (21) Steady industrialized development of geothermal energy in China country update report Proceedings of the World Geothermal Congress 21, Bali, Indonesia, April. 852