Mechanical behaviour, natural permeability, and stimulation of fractured reservoirs Agust Gudmundsson, Ivar Grunnaleite*, Stefan Hoffmann, Belinda Larsen**, Christian Müller, Sonja L. Philipp Department of Structural Geology and Geodynamics Geoscience Centre, University of Göttingen, Germany *The Carbonate Group, International Institute of Research, Stavanger, Norway **Department of Earth Science, University of Bergen, Norway For a natural or man-made geothermal field there must be: a heat source, fluid (water), and sufficient permeability. Example: a high-temperature field in North Iceland. Main high-temperature fields (>200C at 1 km depth) as red dots and low-temperature fields (<150 C at 1 km depth) as brown areas. Volcanic zones (white), volcanic systems (red), and central (composite) volcanoes (black dots). High-temperature fields are mostly confined to central volcanoes (composite volcanoes and calderas), such as the Hengill Volcano close to the capital Reykjavik. A geothermal power plant now about 200 MWe (electricity) but perhaps 300 MWe in the near future. The Hengill Volcano, SW Iceland, is 25 km from Reykjavik. A. T. Gudmundsson, 2001 In a central volcano the main heat sources are swarms of inclined sheets and the shallow magma chamber. Low-temperature fields occur outside central volcanoes; their heat sources are regional dyke swarms and the general heat flow through the crust. Palaeogeothermal fields can be studied in deeply eroded, extinct central volcanoes such as the Geitafell Volcano, Southeast Iceland. S. Burchardt 2006 A narrow depression marks the contact between 100% sheets (right) and the source chamber (gabbro, left) in the 5-6 Ma old Geitafell Volcano, SE Iceland. Gabbro Sheets Part of the Slaufrudalur magma chamber, at 2 km depth. Walls and the roof are well exposed, the latter being dissected by dykes (yellow arrows). Geothermal electricity production in Iceland to 2002: already obsolate figures. Fractured reservoirs can be of various types. In igneous rocks columnar (cooling) joints may generate highly permeable layers. Fingal‘s cave, Scotland Dverghamrar, Iceland Giant‘s Causeway, Ireland (Hatcher Fractures (Bruch) tend to become arrested, or at least offset, at weak contacts between (particularly mechanically dissimilar) layers. Fractured sandstone layers in Germany: Buntsandstein at Reinhausen, Lower Saxony 1m A common feature in layered rocks is that many fractures become arrested at contacts between layers. The percolation threshold may thus not be reached. Limestone (stiff) and shale (soft) layers, Wales, UK About 20 degrees change in trend of joints between limestone layers Mattinata Fault in Southeast Italy is an active strike-slip fault. There are many quarries (carbonate rocks) along the fault trace. Grunnaleite, Carbonate Group, 2006 Whereas other quarries are in the damage zone of the fault. Grunnaleite, Carbonate Group, 2006 To use or make a fractured, fluid-filled reservoir, we must know how its permeability is generated and maintained. Also, how the fractured rock responds to a drillhole, hydraulic fracturing, and/or stimulation? Grunnaleite, Carbonate Group, 2006 Fault Zones, Here Dip-Slip, are Composed Mainly of Two Mechanical Units; a Damage Zone (with Subzones) and a Core Gently Dipping Normal Fault Cutting Limestone and Shale, Kilve Normal Fault, Kilve, England; the main permeability is in the hanging wall. In a subsurface reservoir, a fault contributes to the permeability depending on the infrastructure of the fault, its trend in relation to the current stresses, and its trend in relation to the “hydraulic gradient”. Grunnaleite, Carbonate Group, 2006 Damage zone and core increase in thickness as the fault develops: Normal faults in sandstone and shale, Sinai (Walsh et al. 2002). Most faults have a core and damage zone; in the damage zone the fracture frequency increases towards the core, or the fault plane. Here: a microfault. TaFractures per metre 16 14 12 10 8 6 4 2 0 1 4 7 10 13 16 19 22 25 28 31 34 37 Metre along profile An extinct geothermal system, consisting of mineral veins, in a fault zone in Iceland One can model fluid transport in a fractured reservoir using the program Fred (Fracman); is primarily kinematic. Stumo (Bergen) 2002. Young’s Modulus Decreases Towards the Core of a Fault Core and damage zone tend to have their own local stresses, different from teh surrounding (regional) stress field. Trajectories of σ3. Fault zone with E =1, 5, 10 GPa; host rock E=40 GPa; loading -5MPa Hydraulic fracturing (Smith & Shlyapobersky 2000) Hydraulic fractures are primarily extension fractures, although in detail… (modified from Twiss & Moores 1992) (Brenner 2003) …they are mixed-mode fractures that tend to use the existing weaknesses and fractures in the target layer. Which is of course one reason why we have numerous earthquakes during any type of reservoir stimulation. (Asanuma et al. 2002) Ideally, a hydraulic fracture is a horizontal extension fracture. u W pe Qx 12 x 3 Qx = volumetric fluid flow rate of a water-filled fracture in the horizontal xy-plane For a lateral flow in a fluid-driven fracture along a “neutral buoyancy” layer, the only available driving pressure is the excess pressure in the drill hole. A hydraulic fracture may propagate partly vertically, say along an existing fault in which case buoyancy contributes to the driving pressure. pe W g sin r m 12 L e u QL 3 Qze = Volumetric flow rate through an elastic hydraulic fracture u = Aperture (opening) of the fluid-filled fracture = Dynamic (absolute) viscosity W = Width perpendicular to flow direction m = density of the magma (ρr = density of the host rock) g = acceleration due to gravity sin = dip of the hydraulic fracture pe = excess-pressure gradient in flow direction z Stiff lava flows and soft pyroclastic layers, Iceland Soft Stiff A.T. Gudmundsson, 2001 Hydraulic fracture aperture and the induced stress at the surface depend on mechanical layering. An arrested dyke in the Holocene rift zone, Iceland The mechanical contrast at a contact, rather than layer thickness, decides if a fracture becomes arrested. Shale/limestone contact, Wales. When a fracture meets a multilayer or a weak contact, it commonly becomes arrested (arrows) or, at least, offset. Fracture arrest of this kind is a universal feature of heterogeneous materials, including layered rocks and composite materials, and much studied in fracture, rock, and micromechanics. Carbonate rocks, Mattinata, Italy. Many thin layers A fluid-driven extension fracture follows the trajectories of the maximum principal compressive stress. Stress concentration s3 Stress trajectories s1 Hydraulic fracture tends to open up weak discontinuities such as fault zones. Comsol modelling. Opening depends on attitude of fault in relation to the attitude of the hydraulic fracture. Main reason for opening of discontinuities is fracture-parallel tensile stress. Cook and Gordon, 1964 Beasy model: hydraulic fractures meets a fault zone. Horizontal model Sigma 3 undeformed Host rock 20GPa weak contact 1GPa Overpressure 5MPa Beasy model: hydraulic fractures meets a fault zone. Beasy model: hydraulic fractures meets a fault zone. 60° model undeformed Sigma 3 Host rock 20GPa weak contact 1GPa Overpressure 5MPa Beasy model: hydraulic fractures meets a fault zone. An extension fracture changes to a shear fracture in a soft layer Fault attitude changes from 323/85 in the stiff dyke to 333/65 in the soft pyroclastic rock Hydraulic fracture meeting a weak fault plane tends to… …open up the fault plane and become arrested. So, what do we need to model a fractured fluid-filled geothermal reservoir? Here: Gypsum Veins in Mudstone at Watchet, SW England For example of this type: 1 = injection drillhole, 2 = stimulated fracture system, 3 = production drill holes, and 5-8 = surface installations, e.g. power plant, monitoring and maintenance equipment. (From Geothermische Vereinigung, 2003). Geometry/ Type Geometry/ Type Hydraulics Geomechanics Reservoir Characterisation Geomechanics Heat Flow Answer: We need to know or be able to infer the geometry and types of fractures, its geomechanical behaviour, the fluid transport and heat flow, and the state of stress. How should we make successful man-made geothermal reservoirs? • Make detailed studies of the fault systems, state of stress, and target rocks exposed at the surface, preferably but not necessarily nearby the potential reservoir site. • Combine results on natural geothermal fields (e.g., Iceland) and permeability of subsurface reservoirs (partly studied at the surface; e.g., petroleum companies) to develop permeability models. • Success in use of man-made geothermal energy depends to a large degree on forecasting the potential reservoir-rock fractures, their interconnectivity, and how they respond to stresses (and pressures). • Rock mechanics, fracture mechanics, damage mechanics, micromechanics contain well-developed principles that can help us (with field data) to make accurate forecasts. Results from these fields should be used as principles when making numerical models of fluid-filled fractured reservoirs. Thank you!