Journal of Earth Science, Vol. 26, No. 1, p. 028–036, February 2015 Printed in China DOI: 10.1007/s12583-015-0506-2
ISSN 1674-487X
Recent Development in Numerical Simulation of Enhanced Geothermal Reservoirs Huilin Xing*, Yan Liu, Jinfang Gao, Shaojie Chen Centre for Geoscience Computing (GeoComp), School of Earth Sciences, The University of Queensland, St Lucia QLD 4072, Australia ABSTRACT: This paper briefly introduces the current state in computer modelling of geothermal reservoir system and then focuses on our research efforts in high performance simulation of enhanced geothermal reservoir system. A novel supercomputer simulation tool has been developing towards simulating the highly non-linear coupled geomechanical-fluid flow-thermal systems involving heterogeneously fractured geomaterials at different spatial and temporal scales. It is applied here to simulate and visualise the enhanced geothermal system (EGS), such as (1) visualisation of the microseismic events to monitor and determine where/how the underground rupture proceeds during a hydraulic stimulation, to generate the mesh using the recorded data for determining the domain of the ruptured zone and to evaluate the material parameters (i.e., the permeability) for the further numerical analysis and evaluation of the enhanced geothermal reservoir; (2) converting the available fractured rock image/fracture data as well as the reservoir geological geometry to suitable meshes/grids and further simulating the fluid flow in the complicated fractures involving the detailed description of fracture dimension and geometry by the lattice Boltzmann method and/or finite element method; (3) interacting fault system simulation to determine the relevant complicated rupture process for evaluating the geological setting and the in-situ reservoir properties; (4) coupled thermo-fluid flow analysis of a geothermal reservoir system for an optimised geothermal reservoir design and management. A few of application examples are presented to show its usefulness in simulating the enhanced geothermal reservoir system. KEY WORDS: numerical simulation, geothermal, EGS, microseismicity, finite element method, lattice Boltzmann method. 0
INTRODUCTION Over the past 30 years, a large amount of research and field testing on engineered geothermal system (EGS), also known as hot dry rock (HDR) and hot fractured rock (HFR) geothermal reservoirs, have been accomplished worldwide which include the reservoir construction, fluid circulation and heat extraction (e.g., Tester et al., 2006). A successful EGS reservoir depends on thermal-fluid flow at any given time which primarily determined by its mean temperature and pressure, the nature of the interconnected network of hydraulic stimulated joints and open fractures (including both stimulated and natural), the cumulative amount of fluid circulation (reservoir cooling) that has occurred and water loss (Rybach, 2010; Brown et al., 1999). The reservoir characteristics are complicated and functions of the applied reservoir pressure/stress that are controlling the nature and degree of interconnection within the network of fractures, therefore it is crucial to have good measures and understanding of such reservoir characteristics (i.e., permeability) *Corresponding author:
[email protected] © China University of Geosciences and Springer-Verlag Berlin Heidelberg 2015 Manuscript received March 12, 2014. Manuscript accepted September 26, 2014.
for better understanding, modelling and predicting the thermal power performance of an EGS reservoir. The further literature surveys (e.g., Ingebritsen et al., 2010; Bjornsson and Bodvarsson, 1990) on thermal, hydrological and chemical characteristics of geothermal reservoirs and their relevant parameters— permeability, permeability-thickness, porosity, reservoir temperature and concentration of dissolved solids and non-condensable gases—suggest that reservoir permeability, porosity and total dissolved solids tend to be a function of temperature (Driesner and Heinrich, 2007; Cox et al., 2001). All the above demonstrate that an EGS behaves as multiphysically coupled nonlinear dynamic phenomena in a complicated fractured reservoir system. This requires a more comprehensive understanding and modelling of coupled processes than that commonly done in standard reservoir engineering. Recent studies on computer modelling the conventional geothermal reservoir engineering and the EGS/HDR system are reviewed by O’Sullivan et al. (2001), Sanyal et al. (2000) and Ingebritsen et al. (2010), respectively. More recently, researchers worked into the pre-stimulation coupled THM (thermal hydrological mechanical) modeling (e.g., Rutqvist, 2014; De Simone et al., 2013) and applied in the Northwest Geysers EGS Demonstration Project (Rutqvist, 2014); Held et al. (2014) proposed a workflow on economic evaluation of EGS geothermal reservoir performance through modeling the complexity of the operating EGS
Xing, H. L., Liu, Y., Gao, J. F., et al., 2015. Recent Development in Numerical Simulation of Enhanced Geothermal Reservoirs. Journal of Earth Science, 26(1): 28–36. doi:10.1007/s12583-015-0506-2
Recent Development in Numerical Simulation of Enhanced Geothermal Reservoirs in Soultz-sous-Forêts; Zeng et al. (2014) tried to evaluate the electricity generation potential from fractured granite reservoir at Yangbajing geothermal field through numerical simulations; Jeanne et al. (2014) tried to use a coupled analysis of the microseismicity and the surface deformations to build up a 3D hydrogeological and geomechanical model of the EGS in Geysers geothermal field. They showed that computer modelling is routinely applied in conventional hydrothermal reservoir engineering, but it is comparatively immature in EGS despite lots of related research carried out. Based on the above and other recent studies, existing EGS simulators are faced with following challenges: (a) Geomechanical deformation/rock stress and its full coupling with the multiphase thermal-fluid flow and chemicals are not well addressed yet. Such coupled models are critical for analysing the geothermal reservoir system especially for EGS. Xiong et al. (2013) proposed a fully-coupled fluid flow, geomechanics, and sequentially coupled reactive geochemistry program based on TOUGH2 towards simulating multiphase EGS reservoirs, but the geomechanical model only calculates the mean normal stress instead of stress tensor, and thus cannot analyse the real deformation/fracturing and the related phenomena. Further research is needed in exploring different approximations for coupled processes with vastly different intrinsic spatial and temporal scales for EGS reservoirs. Such a coupled treatment can potentially provide a more realistic description of geothermal reservoir processes during natural/stimulated evolution as well as during exploitation. It can also provide added constraints that can help reduce the inherent uncertainty of geothermal reservoir models. (b) A reliable fully-coupled treatment of 3D fluid flow and mass transport with detailed chemical interactions between aqueous fluids, gases, and primary mineral assemblages still requires further research. This may be currently available in hydrothermal code such as TOUGH2, but not available in the other codes including all the HDR simulators/FEM simulators. (c) The relevant reservoir model generation and meshing are still difficult and time consuming for a fracture dominated geothermal reservoir, especially involving the fracture initiation and propagation during the geothermal energy exploitation. The finite difference method is widely used in geothermal modelling but requires regular rectangular mesh structure. The popular geothermal code TOUGH2 may handle irregular meshes theoretically, but most of models set up using TOUGH2 contain some structure, such as layering. It is impossible to explicitly describe the complicated fractures in an EGS reservoir with such mesh structures. Unstructured mesh may be a better choice, but no unstructured mesh based solver—finite element solver—as powerful as such as TOUGH2 is available for geothermal simulation yet. (d) No module for visualizing microseismicity and evaluating the relevant rupture and permeability distribution for the further simulation has been integrated into such as the finite element based simulator so far. (e) No module for automatic converting the complicated fractures to computational model and further calculating the fluid flow in such complicated fractures considering the detailed fracture size and geometry in an explicit way in geothermal community yet, despite such a computational algorithm (i.e., lattice Boltzmann method) already used in other scientific fields. And (f) no multiscale computing or parallel computing involved in the widely available geothermal simulators yet despite their well-studied and wide-
29
spread application in other fields. In conclusion, further computational model and code developments are urgently needed to improve our understanding of geothermal reservoir and the relevant natural and/or enhanced evolution such as of enhanced geothermal reservoir system, and achieve a more accurate and comprehensive representation of reservoir processes in more details, to reduce the uncertainties in models, and to enhance the practical utility and reliability of reservoir simulation as a basis for field development and management (e.g., Li et al., 2014; Ingebritsen et al., 2010; O’Sullivan et al., 2001, Sanyal et al., 2000). This paper will focus on an introduction of our research efforts towards an integrated high performance simulation of enhanced geothermal reservoir systems. 1 AN INTEGRATED GEOTHERMAL RESERVOIR SIMULATOR Australia has a unique hot fractured rock geothermal resource that could potentially provide enough green energy to meet all its energy needs. A novel finite element and/or lattice Boltzmann method based supercomputer simulation tool PANDAS has been developing for simulating the highly coupled geomechanical-fluid flow-thermal systems involving heterogeneously fractured geomaterials towards address the key scientific and technological challenge in developing EGS energy. Currently, it includes the following six key components: Pandas/Pre, Pandas/Post, Pandas/Thermo, Pandas/Fluid, Pandas/LBM and ESyS_Crustal as detailed in the following (Xing, 2014, 2008; Xing and Zhang, 2009; Xing et al., 2007, 2006; Xing and Mora, 2006; Xing and Makinouchi, 2002a, b, c). (1) Pandas/Pre is developed to (a) visualise the microseismicity events recorded during the hydraulic stimulation process and to further evaluate the fracture location and evolution, geological setting, volume of fractured domain and the relevant material parameters (i.e., permeability) of a certain reservoir; (b) generate the FEM based 2D/3D mesh by using microseismicity data, 2D triangular and quadrilateral mesh by digital imaged structures and complicated fault system (Liu et al., 2011); (c) automatic meshing and construction of a 3D reservoir system through interface with the commercial 3D geological modelling tools (such as GeoModeller and GoCad) which transforms geometric models used in 3D geological visualization into FEM models. Our approach presents a way of using geological data from visualization to construct an FEM mesh for geothermal simulation, which demonstrated by a practical geothermal reservoir system model construction using the practical data provided by Geomodeller and GoCAD (Xing and Liu, 2014; Liu and Xing, 2010). (2) Pandas/post is to visualise the simulation results through the integration of VTK and/or Patran. (3) Pandas/thermo is a finite element method based module for the thermal analysis of the metals and the fractured porous media; the temperature distribution is calculated from the heat transfer induced by the thermal boundary conditions without/with the coupled fluid flow effects in the fractured porous media and the geomechanical energy conversion for the individual/coupled thermal analysis (e.g., Xing, 2014; Xing and Makinouchi, 2002c). It has been also applied in simulating multiple phase coupled flow with phase change (e.g., Bringemeier et al., 2010). (4) Pandas/fluid is a finite element method based module for
30 simulating the Darcy and/or non-Darcy fluid flow in the fractured porous media by solving the conservation equations of macroscopic properties numerically. Here the fluid flow velocity and pressure are calculated from energy equilibrium equations without/with the coupling effects of the thermal and solid rock deformation for the individual/coupled fluid flow analysis (e.g., Liu et al., 2014; Xing, 2014). To simulate the pore scale phenomena in fractured porous media, a lattice Boltzmann method (LBM) based module (Pandas/LBM) is a newly developed for explicitly simulating the 2D/3D complicated fluid flow in the complicated fractures involving the detailed fracture size and geometry at pore scale. Instead of solving the conventional Navier-Stokes equations or its simplified form with macroscopic properties as above in PANDAS/fluid, the discrete Boltzmann equation is solved here to simulate the fluid flow with collision models such as Bhatnagar-Gross-Krook (BGK). LBM models the fluid consisting of fictive particles, and such particles perform consecutive propagation and collision processes over a discrete lattice mesh. By simulating streaming and collision processes across a limited number of particles, the intrinsic particle interactions evince a microcosm of viscous flow behaviour applicable across the greater mass (e.g., Sukop and Thorne, 2007). Due to its particulate nature and local dynamics, LBM has several advantages over other conventional CFD methods, especially in dealing with complex boundaries such as multiphase fluid flow in complicated fractures/porous media with detailed microstructures, incorporating microscopic interactions as well as chemical reactions, and parallelization of the algorithm (Tian et al., 2014; Gao et al., 2013; Gao and Xing 2012a, b). (5) ESyS_Crustal is a finite element method based module developed for the interacting fault system simulation. It employs the adaptive static/dynamic algorithm to simulate the dynamics and evolution of interacting fault systems and processes, in which several dynamic phenomena related with stick-slip instability along the faults need to be taken into account, i.e., (a) slow quasi-static stress accumulation, (b) rapid dynamic rupture, (c) wave propagation and (d) corresponding stress redistribution due to the energy release along the multiple fault boundaries. All these are needed to better describe ruputure/microseimicity/ earthquake related phenomena with applications in earthquake forecasting, reservoir engineering, hazard quantification, exploration, and environmental problems. It has been verified and applied in different cases (e.g., see Xing et al., 2007 and references therein). All the above modules can be used independently/together to simulate individual/coupled phenomena (such as interacting fault system dynamics, heat flow and fluid flow) without/with coupling effects.
Huilin Xing, Yan Liu, Jinfang Gao and Shaojie Chen fracturing in a geothermal reservoir with extremely low permeability. Therefore, finite element method (FEM) based numerical analysis is applied to address such fracturing/faulting issues in an interacting fault system. This includes two major parts: (1) reservoir geological model and mesh generation and (2) numerical analysis by using FEM. The geometrical model construction and mesh generation for such a system is the first step towards numerical simulation through editing and managing the fault/tectonic data and constructing unstructured meshes containing faults/fractures and material boundaries in 2D or 3D. With the advanced imaging and field observation technology, the detailed reservoir geological structure information is available to build computational models. Here, reservoir model construction normally includes the following three aspects: (1) data input, management and editing. Input data may include (a) the reference data for coastlines, the administration field, rivers, the Earth’s surface and the ocean floor; (b) observational data such as the hypocentral distribution and distribution of the distortion; and (c) the underground structural data of stratum boundary point, stratum borderline, stratum and plate/fault, which could be available through the natural and/or artificial seismicity data inversion, drilling well and advanced digital image etc.. Then the input data will be managed and edited for constructing the reservoir model involving the interacting faults, editing and visualising the above data; (2) reservoir geometrical model: construction and editing of the points and curves defining material and fault boundaries, constructing the parametric surfaces using the above curves for further creating and editing of the solid models using the generated surfaces. Finally, output of the solid model of a certain reservoir region will be finalised and outputted for further mesh generation; (3) mesh generation and optimization, specification of loading boundary conditions and material properties etc.. The geometry decomposition is widely used for the hexahedral shaped mesh generation. For a continuous geometry, the actual decomposition of the geometry does not occur, only an internal characterization of sweep/mapped blocks; but for a discontinuous geometry containing faults, the discontinuous boundaries (faults) will be used as a constraint for the geometry decomposition for further mesh generation (Figs. 1 and 2) (Xing and Mora, 2006). The above process may be achieved by numerous in-house and/or commercial software for some cases, but there is so far no software tool available to automatically finalize the above process by generating a high quality hexahedral mesh for a general fractured reservoir system. We are working on towards bridging the gap between geological modelling (such as GoCAD, Petrel and Geomodeller) and numerical simulation as detailed in the Section
2
APPLICATION EXAMPLES PANDAS has been applied to various cases including those in geothermal reservoir systems. A few of relevant applications are briefly reviewed and listed as below. 2.1 Numerical Simulation of Interacting Fault System Dynamics HDR geothermal exploitation faces great challenges from certain geological restrictions, such as faults, high stress and
Figure 1. A 3D fault system model to be analysed.
Recent Development in Numerical Simulation of Enhanced Geothermal Reservoirs
31
Figure 2. A snapshot of the equivalent stress rate distribution simulated by using PANDAS/ESyS_Crustal (the black lines inside are the faults) (Xing and Mora, 2006). 2.4 by using tetrahedron mesh. Numerical modelling of fault systems offers an outstanding opportunity to gain an understanding of their dynamics and complex system behaviour, and to develop the scientific underpinning for risk assessment and management of geothermal exploitation. However, there are a number of challenges that must be overcome. Besides of those discussed as above for a suitable unstructured mesh construction by using geological data, an efficient and stable finite element code is required to simulate the interacting fault dynamics which is more complicated due to the strong nonlinearity related to the stick-slip frictional instability along faults. A finite element module ESyS_Crustal has been developed and tested with the so-called subduction fault model (Xing and Makinouchi, 2002b), sandwich fault model (Xing and Makinouchi, 2002a), single fault bend model and multiple fault bends model (Xing et al., 2006), the Chuan-Dian and Southern California fault model (Xing and Liu, 2011; Xing et al., 2007). Frictional contact problems that are required for fault dynamics simulation are characterized by contact constraints that are imposed on contacting boundaries. Thus, an arbitrarily shaped isoparametric frictional contact element, named as the node-topoint contact element strategy, was proposed and applied here with an efficient contact search algorithm (Xing and Makinouchi, 2002c) in ESyS_Crustal. The static-explicit algorithm (i.e., socalled R-minimum strategy, Xing et al., 2002c) is extended and employed in ESyS_Crustal, which does not involve iterations and hence does not have non-convergence problems. ESyS_Crustal is applied here to simulate stress/velocity variations of a fault system in South Australia in a discrete model with 504 471 nodes constructed using the practical fault data (Fig. 1) supplied by Professor Mike Sandiford of Melbourne University. A snapshot of simulated fracturing and stress variation at a certain stage is shown in Fig. 2. 2.2
Fluid Flow in Fractured Media Both the conventional bounce-back algorithm (Fig. 3) and
the revised force model (Fig. 4) are implemented respectively for impermeable and permeable matrix and compared with each other here. We consider a porous medium with disordered fracture networks as an example (Figs. 3 and 4, with a length of 315 and width of 295 pixels). In both cases, water flow is driven from the bottom to the top under a constant pressure gradient (3e-2Pa/unit) with both left and right sides closed with zero velocity. The differences between the two models are clearly observed in comparison of Figs. 3 and 4, with detailed flux features illustrated inside the porous medium. In the conventional bounce-back LBM model (i.e., impermeable matrix), fluid only flows in some limited pore area and is obstructed at impermeable matrix surfaces (Figs. 3a–3c), while in the revised force model with permeable matrix under the same boundary conditions (Figs. 4a–4c), matrix’s influence on fluid behaviors is obvious: the fluid passes through the permeable matrix and the highest velocity value is less than 1/100 of that in the conventional bounce-back LBM model. From the above comparison, both geometry and its physical characters of the porous matrix have a significant influence on the fluid flow behaviors. 2.3 Pore Scale Simulation of Fluid Flow in 3D Porous Media With the recent development of digital imaging technology, PANDAS is developed to reconstruct the 3D porous medium for discrete LBM lattice core and to simulate the complicated fluid flow at pore scale by using LBM. Scanned CT images of rocks were taken as an application example. The sandstone core is a cube of 2.5 mm in length, and a porous media structure (331×331×331 voxels, Fig. 5a) has been generated according to 600 CT images. It is used here to compare the influence of dissimilar porous matrix on fluid flow behaviour. Two kinds of porous matrix (impermeable matrix and permeable matrix under Weibull distribution with an average porosity of 0.23) are calculated and compared (Figs. 5b and 5c). The
32 flow is driven from bottom to top under a constant dimensionless pressure (0.1 per grid) with all the other sides closed. Although with the same boundary conditions and same geometry, the velocity distributions are clearly different due to dissimilar properties of microscale matrix permeability (Figs. 5b and 5c). Within the impermeable matrix case, velocity distribu-
Huilin Xing, Yan Liu, Jinfang Gao and Shaojie Chen tion differs largely at pores and matrix (Fig. 5b); while within the permeable matrix case, the fluid flows through the matrix with a slower speed than through the fractures (Fig. 5c). It is clearly shown in Fig. 5b that the fluid flow is blocked by the impermeable matrix (districts in black colour), while the seepage phenomenon emerges clearly in the permeable matrix (Fig.
Figure 3. Status of fluid flow passing through the impermeable porous matrix with disordered fracture networks. (a) Velocity profile of fluid flow at time step 100; (b) velocity profile detached from (a); (c) fluid flow route.
Figure 4. Status of fluid flow passing through the permeable porous matrix with disordered fracture networks. (a) Fluid flow route; (b) velocity profile detached from (a); (c) velocity profile of fluid flow at time step 100.
Figure 5. (a) A sandstone sample (331×331×331 voxels) (matrix in grey, pores in black); (b) velocity distribution at time step 1 000 with impermeable matrix (zero velocity in black); and (c) with permeable matrix.
Recent Development in Numerical Simulation of Enhanced Geothermal Reservoirs 5c). It indicates that both porous geometry and property of matrix have significant effects on detailed fluid flow behaviour. 2.4
Automatic Meshing and Model Construction PANDAS/Pre is developed and applied here for automatic meshing and model construction across the different scales of a geothermal reservoir system, which focuses on transforming geometric models and/or digital imaged structures into an FEM mesh in 2D and 3D. A new approach to the automatic generation of an adaptive 2D quadrilateral mesh with arbitrary fault constraints is proposed and developed. It is an indirect all-quad mesh generation including following steps: (1) discretising the constrained lines (fault geometries) within the domain; (2) converting the above domain to an adaptive triangular mesh together with the line constraints; (3) transforming the generated triangular mesh with line constraints to an all-quad mesh through performing an advancing front algorithm from the line constraints, which enables the construction of quadrilaterals layer by layer, and roughly keeps the adaptive feature of the initial triangular mesh; (4) optimizing the topology of the quadrilateral mesh to reduce the number of irregular nodes; (5) smoothing the generated mesh towards highquality all-quad mesh generation. An application example of Chuan-Dian fault system is given to illustrate the meshing process and demonstrate the reliability and usefulness of the proposed algorithm (Liu et al., 2011). It was further developed and applied for meshing multi-material structures with a boundary
33
focused quadrilateral mesh generation algorithm (Liu and Xing, 2013; Xing and Liu, 2011). PANDAS/Pre is also developed and applied for automatic meshing and construction of a 3D reservoir system. The development of digital 3D geological modelling tools and applications (such as GeoModeller, Petrel and GoCad) makes it possible to establish geometric models for 3D geological objects, where the geometric features could be illustrated by a set of triangles. PANDAS/Pre is developed to focus on constructing a reservoir system which transforms geometric models used in 3D geological visualization into FEM models. A 3D Geological model for visualization is used as input here, and the transformation processes are implemented following: (a) abstracting and smoothing the boundaries of different geological objects; (2) tetrahedral mesh generation with the constraints of the refined boundaries; (3) revising the meshed model locally according to users’ requirements, such as setting up well models by tetrahedral mesh refinement. Both geothermal reservoir model constructed by using Geomodeller and Lawn Hill reservoir model constructed by using GoCAD are taken as practical application examples to be analysed (Liu and Xing, 2010). Lawn Hill reservoir model covers 140.5×161.5×19.4 km3 and described in 3D digital image with 71 million voxels, which was used as out input data for mesh generation (Fig. 6). Our approach presents a way of using geological data for visualization to construct an FEM mesh for geothermal simulation, which demonstrated by a practical reservoir system model
Figure 6. 3D mesh generation for the reservoir model. (a) The whole model to be analysed; (b) part of surface boundary extracted and surface mesh generated; (c) volume mesh generated for the bottom strata; (d) a close-up of the volume meshes.
Huilin Xing, Yan Liu, Jinfang Gao and Shaojie Chen
34 construction using the practical data (Figs. 6a–6d). 2.5
Microseismicity and EGS Reservoir Microseismicity is widely used in the mining industry including the hot dry/fractured rocks (HDR/HFR) geothermal exploitation to monitor and determine where/how the underground rupture proceeds. Hydraulic stimulation is the basic concept of improving the residual permeability of the in-situ rock mass at depth and still remains the main mechanism to be envisaged for the creation of an enhanced HDR/HFR/HWR reservoir. For a hydraulic stimulation process, hundreds and thousands of microsesimic events are recorded. The recorded microseismicity data provides the detailed location and time where and when an underground dynamic rupture occurs. We take the recorded data of every event location and its occurrence time as an independent node with its location in 3D space as recorded and colour it with its occurrence time, thus where the underground dynamic rupture locates and how it proceeds with the time during a hydraulic stimulation process will be identified from such nodal information. PANDAS/Pre has been developed and applied to visualize the microseismic events, to monitor and determine where/how the underground rupture proceeds during a hydraulic stimulation process, to determine the domain of the ruptured zone and to evaluate the material parameters (i.e., the permeability) for the further numerical analysis (Fig. 7). Moreover, to determine the maximum domain of the rup-
tured zone for the further analysis through the recorded microseimicity data, the Delaunay algorithm based mesh generation method is developed and applied here. Firstly, the Delaunay algorithm is used to form the convex hull using the above recorded data set, and then the outside surface of the convex hull is extracted. Finally, the Delaunay algorithm is applied to generate the tetrahedral shaped meshes with a certain internal node distribution (such as the variable mesh size control according to the outside surface). The micro-seismicity data visualisation and the generated meshes with the above data at the different hydraulic stimulation stages are shown in Figs. 7 and 8. These clearly show the rock mass ruptured with the microseismic events proceeding and can also be used for the further finite element analysis. A virtual 8-well geothermal reservoir with the dimensions of length×width×height: 4 000 m×3 000 m× 1 750 m) is designed with 7 production wells and 1 injection well and further analysed with the permeability distribution calculated thorough the above microseismic events by using PANDAS. The snapshots of the relevant results of fluid flow, hydraulic pressure variation and temperature distribution at the different stages are shown in Figs. 9a–9c, respectively. 3
CONCLUSIONS Based on the relevant surveys on geothermal reservoirs and the related computer modelling, the key improvements required in simulating an enhanced HDR/HFR/HWR geothermal reservoir system are discussed. Our research in an
Figure 7. Visualization of the micro-seismicity data in a hydraulic stimulation process recorded during the Harbone #1 hydraulic stimulation process by the Geodynamics Ltd. at the different time (days). (a) 16.417 and (b) 37.25 (Wyborn, 2013).
Figure 8. Visualization of the solid domain of the rupture zone occupied in a hydraulic stimulation process at the different time (days). (a) 16.417, and (b) 37.25.
Recent Development in Numerical Simulation of Enhanced Geothermal Reservoirs
35
ating the material parameters (i.e., the permeability), and further in a virtual design and assessment of a multiple well reservoir system based on permeability distribution calculated from the recorded microseismic events. The related application examples demonstrated its stability and potential usefulness in simulating the enhanced geothermal reservoir systems. ACKNOWLEDGMENTS Support was gratefully acknowledged by the ARC (Nos. LP0560932, DP110103024), the NSFC (Nos. 51034003 and 51174210) as well as Geodynamics Limited. The authors are grateful to Drs Doone Wyborn, Hehua Xu, Wenhui Yu, Ji Zhang, Zhiwei Tian, Enlong Liu and Profs. Peter Mora and Hans Muhlhaus, for their helpful discussions in the relevant research.
Figure 9. The simulated results in a certain fractured geothermal reservoir with 7 production wells and 1 injection well H1. (a) Fluid flow shown in a certain cross-section; (b) hydraulic pressure distribution at the time of 5 years and (c) temperature distribution at the time of 40 years. integrated geothermal reservoir simulator PANDAS for high performance simulation of enhanced geothermal reservoirs is introduced and then tested by the relevant cases. PANDAS has been successfully applied in the related geothermal model construction and analysed by using finite element and lattice Boltzmann method, and further applied in visualising the recorded microseismic events, monitoring and determining where/how the underground rupture proceeds during a hydraulic stimulation process, determining the domain of the ruptured zone and evalu-
REFERENCES CITED Bjornsson, G., Bodvarsson, G., 1990. A Survey of Geothermal Reservoir Properties. Geothermics, 19: 17–27 Bringemeier, D., Wang, X., Xing, H. L., et al., 2010. Modelling of Multiphase Fluid Flow for an Open Pit Development within a Geothermal Active Caldera. In: Proceedings of the 11th IAEG Congress (International Association for Engineering Geology and the Environment). Auckland, New Zealand Brown, D., DuTeaux, R., Kruger, P., et al., 1999. Fluid Circulation and Heat Extraction from Geothermal Reservoirs. Geothermics, 28: 553–572 Cox, S. F., Knackstedt, M. A., Braun, J., 2001. Principles of Structural Control on Permeability and Fluid Flow in Hydrothermal Systems. Reviews in Economic Geology, 14: 1–24 De Simone, S., Vilarrasa, V., Carrera, J., et al., 2013. Thermal Coupling may Control Mechanical Stability of Geothermal Reservoirs during Cold Water Injection. Physics and Chemistry of the Earth, Parts A/B/C, 64: 117–126 Driesner, T., Heinrich, C. A., 2007. The System H2O-NaCl. I. Correlation Formulae for Phase Relations in TemperaturePressure-Composition Space from 0 to 1 000 °C, 0 to 5 000 bar, and 0 to 1 XNaCl. Geochimica et Cosmochimica Acta, 71: 4880–4901 Gao, J., Xing, H. L., 2012a. LBM Simulation of Fluid Flow in Fractured Porous Media with Permeable Matrix, Theor. Appl. Mech. Lett., 2: 032001(4) Gao, J., Xing, H. L., 2012b. High Performance Simulation of Complicated Fluid Flow in 3D Fractured Porous Media Using LBM. In: 10th International Meeting on HighPerformance Computing for Computational Science. Kobe, Japan Gao, J., Xing, H. L., Tian, Z., 2013. Lattice Boltzmann Modeling and Evaluation of Fluid Flow in Heterogeneous Porous Media Involving Multiple Matrix Constituents. Computers & Geosciences, 62: 198–207 Held, S., Genter, A., Kohl, T., et al., 2014. Economic Evaluation of Geothermal Reservoir Performance through Modeling the Complexity of the Operating EGS in Soultz-Sous-Forêts. Geothermics, 51: 270–280 Ingebritsen, S. E., Geiger, S., Hurwitz, S., et al., 2010. Numerical Simulation of Magmatic Hydrothermal Systems. Rev. Geo-
36 phys., 48: RG1002. doi:10.1029/2009RG000287 Jeanne, P., Rutqvist, J., Vasco, D., et al., 2014. Development of a 3D Hydrogeological and Geomechanical Model of an Enhanced Geothermal System Using Microseismic and Ground Deformation Data from a 1-year Injection Program. In: Proceedings, Thirty-Ninth Workshop on Geothermal Reservoir Engineering Stanford University. Stanford, California Li, Q., Wei, Y. N., Liu, G., et al., 2014. Combination of CO2 Geological Storage with Deep Saline Water Recovery in Western China: Insights from Numerical Analyses. Applied Energy, 116: 101–110 Liu, X., Xing, H. L., Zhang, D., 2014. Fluid Focusing and Its Link to Vertical Morphological Zonation at the Dajishan Vein-Type Tungsten Deposit, South China. Ore Geology Reviews, 62: 245–258 Liu, Y. H., Xing, H. L., Guan, Z. Q., 2011. An Indirect Approach for Automatic Generation of Quadrilateral Meshes with Arbitrary Line Constraints. International Journal for Numerical Methods in Engineering, 87(9): 906–922. doi:10.1002/nme.3145 Liu, Y., Xing, H. L., 2010. Automatic Meshing and Construction of a 3D Reservoir System: From Visualization towards Simulation. In: Gurgenci, H., Weber, R. D., eds., Proceedings of the 2010 Australian Geothermal Energy Conference. Nov. 17–19, Adelaide Liu, Y., Xing, H. L., 2013. A Boundary Focused Quadrilateral Mesh Generation Algorithm for Multi-Material Structures. Journal of Computational Physics, 232: 516–528 O’Sullivan, M. J., Pruess, K., Lippmann, M. J., 2001. State of the Art of Geothermal Reservoir Simulation. Geothermics, 30: 395–429 Rutqvist, J., 2014. Pre-Stimulation Coupled THM Modeling Related to the Northwest Geysers EGS Demonstration Project. In: 38th Annual Workshop on Geothermal Reservoir Engineering, Stanford Rybach, L., 2010. The Future of Geothermal Energy and Its Challenges. In: Proceedings World Geothermal Congress 2010. 25–29 April, 2010, Bali Sanyal, S. K., Butler, S. J., Swenson, D., et al., 2000. Review of the State-of-the-Art of Numerical Simulation of Enhanced Geothermal Systems. In: Proceedings World Geothermal Congress 2000, Tohoku. 3853–3858 Sukop, M. C., Thorne, D. T., 2007. Lattice Boltzmann Modeling: An Introduction for Geoscientists and Engineers. Springer Tester, J. W., Anderson, B. J., Batchelor, A. S., et al., 2006. The Future of Geothermal Energy—Impact of Enhanced Geothermal Systems (EGS) on the United States in the 21st Century. Massachusetts Institute of Technology, Cambridge. 358 Tian, Z., Xing, H. L., Tan, Y., et al., 2014. A Coupled Lattice Boltzmann Model for Simulating Reactive Transport in CO2 Injection. Physica A: Statistical Mechanics and Its Applications, 403: 155–164 Wyborn, D., 2013. The Innamincka EGS Project—10 Years of Operation. http://cug60.cug.edu.cn/dgs2012/ 11%20Doone%20Wyborn.pdf
Huilin Xing, Yan Liu, Jinfang Gao and Shaojie Chen Xing, H. L., 2008. Progress Report: Supercomputer Simulation of Hot Fractured Rock Geothermal Reservoir Systems. ESSCC/ACcESS Technical Report, The University of Queensland. 1–48 Xing, H. L., 2014. Finite Element Simulation of Transient Geothermal Flow in Extremely Heterogeneous Fractured Porous Media. Journal of Geochemical Exploration, 144(Part A): 168–178. doi:10.1016/j.gexplo.2014.03.002 Xing, H. L., Liu, Y., 2011. Automated Quadrilateral Mesh Generation for Digital Image Structure. Theoretical and Applied Mechanics Letters, 16: 061001.1-061001.3. doi:10.1063/2.1106101 Xing, H. L., Liu, Y., 2014. Mesh Generation for 3D Geological Reservoirs with Arbitrary Stratigraphic Surface Constraints. Procedia Computer Science, 29: 897–909. doi:10.1016/j.procs.2014.05.081 Xing, H. L., Makinouchi, A., 2002a. Finite Element Analysis of Sandwich Friction Experimental Model of Rocks. Pure and Applied Geophysics, 159: 1985–2009 Xing, H. L., Makinouchi, A., 2002b. Finite-Element Modelling of Multibody Contact and Its Application to Active Faults. Concurrency and Computation: Practice and Experience, 14: 431–450 Xing, H. L., Makinouchi, A., 2002c. Three Dimensional Finite Element Modelling of Thermomechanical Frictional Contact between Finite Deformation Bodies Using R-Minimum Strategy. Computer Methods in Applied Mechanics and Engineering, 191: 4193–4214 Xing, H. L., Makinouchi, A., Mora, P., 2007. Finite Element Modelling of Interacting Fault System. Physics of the Earth and Planetary Interiors, 163: 106–121 Xing, H. L., Mora, P., 2006. Construction of an Intraplate Fault System Model of South Australia, and Simulation Tool for the iSERVO Institute Seed Project. Pure and Applied Geophysics, 163: 2297–2316. doi:10.1007/s00024-006-0127-x Xing, H. L., Mora, P., Makinouchi, A., 2006. A Unified Frictional Description and It’s Application to the Simulation of Frictional Instability Using the Finite Element Method. Philosophical Magazine, 86(21/22): 3453–3475 Xing, H. L., Yu, W., Zhang, J., 2009. 3D Mesh Generation in Geocomputing. In: Xing, H. L., ed., Advances in Geocomputing. Springer-Verlag, Berlin & Heidelberg. 27– 64. doi:10.1007/978-3-540-85879-9_2 Xing, H. L., Zhang, J., 2009. Finite Element Modelling of Nonlinear Deformation of Rate-Dependent Materials Using a R-Minimum Strategy. Acta Geotechnica. 4(2): 139–148. doi:10.1007/s11440-009-0090-7 Xiong, Y., Fakcharoenphol, P., Winterfeld, P., et al., 2013. Coupled Geomechanical and Reactive Geochemical Model for Fluid and Heat Flow: Application for Enhanced Geothermal Reservoir. In: SPE Reservoir Characterization and Simulation Conference and Exhibition, Society of Petroleum Engineers Zeng, Y. C., Wu, N. Y., Su, Z., et al., 2014. Numerical Simulation of Electricity Generation Potential from Fractured Granite Reservoir through a Single Horizontal Well at Yangbajing Geothermal Field. Energy, 65: 472–487
