Click for next page ( 2


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
THE ROLE OF FLUIDS IN CRUSTAL PROCESSES

OCR for page 1

OCR for page 1
Overview and Recommendations SUMMARY The central role of H2O-rich fluids in determining the dynamic conditions in the Earth's crust is apparent in the repeated occurrence of fluid properties in all of the transport equations. This symbolic depiction of the processes shows not only the influence of processes on one another but also that this coupling condition is a consequence of the presence of an often sparse, but essential, occurrence of water in the systems. Numerous examples exist that demonstrate water as an active agent of the mechanical, chemical, and thermal processes that control many geologic processes that operate within the crust. This study assesses the current scientific understanding of the role of fluids in crustal processes. An important part of such an assessment is an evaluation of the adequacy of the geophysical knowledge base and the opportunities to improve it. Recognition of the role of water as the material that controls the extent of coupling among processes shows promise of reducing the magnitude of the analytical problem to one that focuses on the controlling links in the system. INTRODUCTION Fluids play a vital role in virtually all crustal processes. Circulation of fluids has important effects on the transport of chemical constituents and heat, and is the principal control on the formation of hydrothermal ore deposits. The mechanisms by which crustal rocks deform are strongly influenced by the presence of water, as well as by the pore-fluid pressure. It has also been suggested that fluids play a very significant role in earthquakes. On a broader scale, pore-fluid pressure influences the mechanical processes that control rock deformation in and below the accretionary wedge in subduction zones. The volume of fluid carried to deeper levels in subduction zone complexes appears to influence the rate and depth of melting, which will determine the site of volcanism in the overriding plate. 3

OCR for page 1
4 OVERVIEW AND RECOMMENDATIONS (See Vrolijk and Myers, Chapter 10, for additional discussion.) Development of rigorous models of crustal dynamics, therefore, require a substantial body of well constrained data as well as in-depth understanding regarding fluids in the crust. Geologists have traditionally focused on the solids, minerals, and lithologic units in their study of earth processes; the importance and distribution of the fluid phase often has been overlooked. This is in part due to the difficulty of obtaining samples of the fluids. Investigations that have not taken into account the presence of water in the Earth may inadequately describe geophysical mechanisms in the Earth or describe them incorrectly. Hydrologists and reservoir engineers approach problems associated with the Earth differently than geologists. With hydrologists, the focus is on the fluid. The questions are: What is the current state of fluid; how is the fluid moving; how is it transporting mass and energy? Implicit in this approach is the concept that if one can understand how the fluids move and transport mass and energy, then one can better understand the dynamics of geologic processes within the crust. Once rock has formed in the Earth's crust, moving fluids and material transport are most likely to produce change in the rocks. When geologists treat problems of fluid-rock interaction they have generally done so from the perspective of geochemistry (see for example, Fyfe et al., 1978~. The geochemi- cal approach lacks information concerning the motion of the fluid. In assembling this report we emphasized formulating the problem in terms of mass and energy transport, an approach rather different from classical geochemistry. Our approach resembles non- equilibrium thermodynamics. To many trained in geochemistry, this approach appears to neglect the chemistry. Problems of subsurface fluid flow appear to be complex. However, the physics and chemistry are described by a set of conservation statements for mass, energy, and momen- tum, which leads to a set of coupled partial differential equations. The dependent variables of interest are fluid pressure, chemical composition of the fluid, and fluid temperature or enthalpy. The coupled equations must be solved simultaneously. The equations are difficult to solve analytically; however, with modern digital computers, solutions to realistic problems of great scientific interest are now readily possible. While the results of such a dynamic system may be complex, in principal, the physics and chemistry are conceptually simple. The characteristics of flowing fluid systems are determined by the bulk permeability, pressure gradient, fluid composition, temperature, and available fluid volume. At shallow crustal levels (depths less than several kilometers) these variables are measured directly in wells. However, for rocks at deeper crustal levels, the bulk of available information regarding the behavior of fluids comes from observation of exposed rocks that once resided at deeper crustal levels. These outcrops provide a series of "snapshots" (often multiple exposures) that, when integrated, allow inferences to be drawn concerning the behavior of fluids at different levels of the crust; the magnitude of the relevant variables noted above can only be established within very broad limits. Since fluids present when the strata were at deeper crustal levels have long since escaped (or at least changed) from the rock system, these "snapshots" have the disadvantage of providing only a record of the effect fluids had on the solid crust, rather than providing a sample of the fluid and its role in crustal development. A problem of using fluid inclusions is not so much in interpreting a "snapshot," but in deciding which one, of numerous multiple exposures, is appropriate for the fluid sample preserved. Present day surface exposures indicate that, at all crustal levels, fluids have been present. Field, isotopic, fluid inclusion, and phase equilibria studies of rocks and mineral- ized fractures from shallow crustal levels, vein and pegmatitic masses from intermediate levels, and gneissic units from deeper crustal levels, document that fluids were present in significant volumes. These investigations (see e.g., Wickham and Taylor, Chapter 6) have demonstrated that rocks may be chemically modified as a result of fluid migration. Consis- tent with this fact is the repeated observation that fluids have played a dominant role in the transport and concentration of metals in most ore deposits.

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 5 Geophysical studies have also shed light on the nature of deep crustal fluids. Electro- magnetic and conductivity soundings have revealed zones of relatively low electrical resistivity in the deep crust, suggesting the presence of a continuous aqueous fluid phase. Seismic reflections and inferred low velocity zones have been interpreted to be regions in which the fluid pressure is anomalously high, in some cases approaching the lithostatic pressure (see Oliver, Chapter 8, this volume). Table 1 summarizes much of the observa- tional geochemical and geophysical indicators that bear on the question of the extent of free water in the Earth's crust. In addition to these indicators, fluids have been directly sampled at about 11 km by the Soviets at the Kola Peninsula drillhole. These data collectively support the existence of fluid circulation and therefore fractures to crustal depths of at least 10 to 15 km and perhaps significantly deeper. Some evidence exists that nonaqueous fluids (e.g., CO2 and gaseous and liquid hydrocarbons) may also be present. Despite these results, there is very little quantitative information from depths greater than 5 km regarding the actual processes that lead to fluid migration, heat and mass transport, the development of mineral deposits, and the role of fluids in earthquake events. We remain at the stage in which each new field area provides a relatively unique picture of the processes associated with fluids in the crust. There is an obvious need to sample the deeper crust through drilling. On the basis of the mechanical properties of crust, we can infer that, locally, large relative volumes of fluid may be expected at shallow crustal depths, and vanishingly small volumes may be present at deep crustal levels. This conclusion reflects the effects of brittle versus ductile behavior of rocks, their pore size and permeability reduction at higher pressures, and the local presence or absence of fluid sources. However, we remain ignorant of how the transition from high fluid volume regimes to low fluid volume regimes occurs. Since the crust must be considered a dynamic system in which rocks at a given crustal depth may be forced to shallower or deeper depths through time, it is also important to understand how the transition from one regime to another can influence the chemical and physical properties of any given rock through time. In addition, we do not know the actual range of volumes to be expected at any particular crustal depth; some studies indicate that some chemical modification of deep crust (which requires significant fluid volumes) has occurred, contrary to inferences based solely on inferred mechanical properties. One example of how H2O-rich fluid strongly affects the evolution of conditions in the crust is the central role it plays in hydrothermal systems. At constant volume, a rise in temperature of 1C can cause a pressure increase of several bars in the local fluid pressure. TABLE 1 Possible Indicators for the Presence of Free Water in the Earth's Crust as Suggested by Various Investigators (See Table 7.1, Chapter 7, for complete references.) Indicator Water table Deep wells Reservoir induced seismicity Crustal low velocity zones Crustal electrical conductivity zones Oxygen isotopes Metamorphism Crack healing and sealing Formation of hydrothermal ore deposits Crustal seismic attenuation zones Low stress on faults Silicic volcanism Fluid inclusions Depth Range 0 to 2 km to 12 km to 12km 7 to 12 km 10 to 12 km to 12 km to >20 km . . . to >5 km 7 to 12 km 0 to 10 km near surface . . .

OCR for page 1
6 CONDUCTIVE Ed CONVECTIVE COOLING RATE OVERVIEW ANrD RECOMMENDATIONS THERMS ) PERT - ATI~ ~ ,~ - \ FIGURE 1 Systematic relationships between prin- cipal transport rates and their products. Arrows depict directions of energy, mass flow, and feed- back effects of state conditions on rates (from Norton, 1984~. AUCTION RATE \ CRYSTALLIZATION ~R; STATE OF SASSY f RACTURE _ ~- d*lA-~^ A ~__ STRAIN RATE - ADUI'UAI~C ~ ~ A~RTURE CHEMICAL AfFlHITY - CHEMICAL REACTION RATE The introduction of a large body of magma into the crust can result in a convecting groundwater system around the magma body, which transports heat away from the magma at supercritical conditions. Because a supercritical fluid is efficient as a heat-transporting medium, the rates of groundwater flow will control the cooling rates of the igneous body. The same convective process will also redistribute chemical components from the pluton to the host environment, including ore-forming constituents. Groundwater flow controls the distribution and grade of ore deposits. The magnitude of groundwater convection and consequently the extent of mineral alteration around a magma body depends on the permeability of rocks. However, even if the rocks were initially impermeable, the pore-fluid pressure generated by heat dispersed into the host rocks is high enough to create hydraulic fractures. Thus, the system can generate its own permeability by making fractures. This fracturing process tends to be episodic. Repeated sequences of fracturing followed by mineral deposition in the fractures are typical of magma environments. As the fractures gradually fill with minerals, permea- bility decreases, and the fluid flow is retarded. This cycle of processes forms an intricately interconnected feedback system (Figure 1) in which the properties of the fluid phase exert primary control on the evolution of the system. CONSERVATION EQUATIONS Physical Parameters The concept of conservation provides a basis for writing a set of equations that symboli- cally represent the transport processes involving water in the Earth's crust. For mass, energy, and momentum, equations are formulated that express the conservation of that quantity with respect to the local system. These equations (see Chapter 1, for details) describe the rate of change in these quantities with respect to time in a representative rock volume of the system. During the nineteenth century some carefully conducted experiments revealed a set of

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 7 empirical laws that express the flux of mass and energy in terms of a driving force and a proportional constant that incorporates a medium's properties. These flux laws form the basis for discussing transport; they include Fick's Law of Diffusion, Fourier's Law of Heat Conduction, DeDonder's Law of Affinity, and Darcy's Law of Fluid Flow. Each has the general form of the flux of a particular quantity, where the flux is proportional to the gradient in a field parameter. The conservation equations for mass, energy, and momentum all derive from one or more of these flux laws. The set of appropriate differential equations to understand the physical processes is summarized in Table 2. Table 2 is set up (column 1) assuming no coupling; in other words, each state of fluid-pressure, composition, tem- perature, and so on, can be derived independently. This is done purely for simplicity; in general, everything is coupled. Table 2 illustrates how much information is required to solve any one problem of interest. In each case the movement of mass and energy involves an empirical relationship and an empirically derived parameter that describes a property of the medium. For example, in groundwater the flow of fluid through a porous medium is described by Darcy's Law, which is an empirical relationship requiring an empirically derived parameter the "per- meability" of the porous medium. For each problem there is a set of such parameters that describes the physical chemical setting of the problem being considered. Table 3 lists in more detail the parameters that are required to solve a realistic fluid transport problem. Many of the parameters are both scale and time dependent. The TABLE 2 Information Required to Address the Major Vanables Involved with Understanding Fluid Properties (These individual properties need to be coupled in order to assess the role of fluids in crustal processes.) System State Information Required Output Pressure (isothermal Geometry Pore fluid and no change in Boundary Conditions pressure fluid chemistry) Parameters Fluid velocity Hydraulic conductivity distribution Porosity Elastic constants Sources and sinks of fluid Mass or Chemical Fluid velocity distribution Composition of Transport (steady Chemical boundary conditions fluids and rocks flow and isothermal) Parameters Porosity Dispersivity Chemical Reactions Identify Kinetics Heat Transport Fluid velocity distribution Temperature (steady flow and Heat boundary conditions no change in fluid Parameters chemistry) Porosity Thermal conductivity Thermal dispersivity Enthalpies Stress and Strain State of stress Strain and failure (elastic) Fluid pressure (fracture) Temperature Parameters Elastic constants

OCR for page 1
8 OVERVIEW AND RECOMMENDATIONS number of x's are a subjective attempt to indicate where the parameters can be investigated, either in the field or the laboratory, and which are most important in obtaining a realistic solution. Two parameters and their changes during flow turn out to be critical in the transport process: (1) the permeability and (2) the chemical reactions and their kinetics. The pathways for flow and therefore the nature of permeability is critical to any study of fluid flow and transport. What are the flow pathways that exist in the crust? Brace (1980) summarized both the laboratory and in situ large scale permeability data for crystalline rocks (Figure 2~. He concluded that the laboratory permeability of most of these rocks, measured on small samples is usually several orders of magnitude less than the in situ permeability measured at a larger scale in the Earth's crust. Bredehoeft et al. (1983) made a similar observation for the Cretaceous Pierre Shale in South Dakota. This differ- ence presumably reflects the dominant effect of fluid flow along widely spaced fractures in the crust. As fractures become less prominent at deeper crustal levels where pressures and temperatures are higher, the mechanism of dominant flow becomes ambiguous. Whether fluids move via grain boundary migration, fluid overpressure/hydrofractur~ng, or diffusion along defects remains unknown. Changes in grain boundary geometry occur when fluid compositions change, resulting in significant variation in permeability and porosity. The extent to which such changes are important in controlling fluid migration is unclear. Migrating fluids can encourage recrystallization of the rocks through which they pass; if this occurs, changes in grain geometry can lead to variations in the local porosity. Devel TABLE 3 Relative Need of System Properties Important to Understanding Coupled Fluid Properties and Dynamics (the greater the need, the greater number of x's) Scale Effects System Properties Lab Field Time Dependent Effects Lab Field Basic Geometry xxxx (structure and stratigraphy) Boundary Conditions Topography xxxx Heat Sources xxxxxx Tectonic Effects (rock mechanics) xx xx xx Sources and Sinks - xxx xxx Parameter Distribution Permeability Porous media xxxxxx xxx Fracture media xxxxx xxx Porosity xxxx xx Fluid Density x x Viscosity xx x Chemical Reactions Identification xxxxxx Kinetics xxxxx Dispersivity Mass xxx x Heat xx x Thermal Conductivity xxx x Elastic Constants xxx Plastic Behavior xxxxx Note: Also shown is the type of investigations needed relative to the field or laboratory.

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 9 Rocks | Unconsolidoted k 11 C- o o _ In ~ .~ .E ',, I _ ~ our y E, o 1 1.2-~1 1 o o ~ ~ =. O E ~ ' ~ ~` 1 ~ ~ ~ ~ . O 11 FIGURE 2 Range of in situ permeability measurements for crystalline and argillaceous rocks from various sites around the world (after Brace, 19801. ~ - =! .O ~ ^~ . ~ coo ~c' ~ ~ G ~r._ ~ ~ O ~ E .=,.4 1 . . ~1 0 _ _ ._ en ~ K ~ ~ (cm2) (cm/s) (m/s) (gal/doy/ft2) 105 10-] 102 -104 -10-4 -10 -103 _lo-5 -1 - 1 o2 - 1o-6 - 1 o-l -10 -10 ~_lo-2 - t - 10-8 - 10-3 -10 ~-10-9 -10 - - 1 o~2 - 1 o-Io - 1 o-5 - 1 0~3 - 10-11 - lo~6 - 10-4 - lo~12 - 10-7 t0-S -10-13 -10-8 - to-6 - 10-14 - 10-9 - I 0~7 - to-ts - lO-10 - 10 ~-10-t6 -10 ', rl 10-1 lo-2 10-3 lo-4 - lo-s - lo-6 -1~7 -10-8 - 10-9 lo-lo -10 lo-2 lo-13 - l o6 - 10 - 104 log - lo2 - 10 1 - 10-1 -,o-2 - 10 lo-4 ,o-s ,o-6 - 10-7 opment of preferential, or channelized, flow regimes may occur where local conditions or processes enhance permeability. The extent to which porous media versus channelized flow occurs has yet to be determined. Mineral-Fluid Reactions The advection of chemical components by fluid flow and dispersive fluxes from one chemical environment to another causes chemical reactions between the minerals and fluids. The processes of dissolution, precipitation, ion exchange, and sorption can all be represented by the general equations for the conservation of mass. The general problem of chemical mass transport by groundwater is one of considering advective mass transport through a region in which reversible chemical reactions must be considered in terms of overall disequilibrium, yet with local equilibrium among some of the minerals and the aqueous phase. This condition has long been recognized in the study of natural weathering and hydrothermal processes and more recently in engineering studies of contaminant transport in groundwater. Although the analysis of chemical contaminants in shallow groundwater systems does not involve the long periods typical for geologic problems, both situations rely on a similar formulation. Equilibrium is simply a special case in the more general formulation. The irreversible formulation requires that we understand the kinetics of the reactions of interest. The equilibrium between minerals and the fluid require a set of thermodynamic relations to describe their activity-composition relationships, and the minerals out of equi- librium require a kinetic-rate law consistent with the thermodynamic standard state to describe their rate of change. Also, a single equation must be written for each of the basic chemical components. In typical problems related to ore deposition, from 15 to 20 equations have nonlinear features that make them difficult to integrate numerically; realis- tic problems require large computer capacity, often supercomputers. For contaminant transport problems, as many as five simultaneous concentration equations have been used to analyze specific problems.

OCR for page 1
10 OVERVIEW AND RECOMMENDATIONS COUPLING How are thermal, mechanical, hydrological, and chemical processes coupled? The closed nature of many mineralized fractures illustrates the important interaction between the chemistry and physical transport properties of the system. In general, fluid flow transports both heat and solute mass. This transport can perturb a system in chemical equilibrium into a disequilibrium state, and consequently, chemical reactions between the fluid and rock will occur. These reactions dissolve and precipitate solid material, changing the hydraulic characteristics of the fluid pathways and inducing heterogeneities in perrnea- bility and dispersivities; these changes in turn modify fluid flow patterns and mechanical properties. In multiphase fluid systems, heat transfer can also modify in situ fluid satura- tions through boiling, evaporation, and condensation. Saturation changes can also modify transport properties by changing relative permeabilities. Chemical reactions within the system are strongly dependent on the salvation properties of the aqueous phase, because this phase controls the dissolution and deposition of miner- als, and hence the porosity distribution. Diffusion and advection processes are controlled by porosity and permeability. Ultimately, fluid transport is coupled to solution and precipi- tation. If rock strain is considered to be dependent on changes in total stress, as well as temperature, then pore pressure is strongly coupled to heat transport. In some contexts it is possible to ignore the coupling and thus simplify the appropriate mathematics. The equations for pressure, chemical composition, temperature, and Darcy's Law are coupled and nonlinear. An additional coupling occurs because of the dependence of both fluid density and viscosity on pressure, temperature, and concentration. The propor- tionality constant in Darcy's Law includes properties of the fluid, both density and viscos- ity. Density is influenced by pressure, temperature, and chemical concentration, whereas for most problems, viscosity is strongly influenced only by the temperature. To fully understand the role of fluids in crustal processes, it will be necessary to unravel the complex coupling between thermal, chemical, mechanical, and hydrological processes. Each of the processes, properties, and driving forces that are currently included in the basic theory is the consequence of existing experimental data, theoretical analysis, and computer simulation. Few of these linkages have been directly observed in the field, yet it is likely that it is the interaction of these processes that ultimately controls, to a large extent, the movement and transport by fluids in the Earth's crust. Understanding these complex coupled processes is not only necessary for unraveling the nature and origin of mineral deposits, hydrothermal systems, and fluid-dependent crustal changes, but it is also needed to rationally address many societal issues, including groundwater contamination, radioac- tive and hazardous waste disposal, enhanced oil recovery, and exploitation of geothermal heat. Obtaining clearer answers to these questions will provide the means to establish the extent to which fluids have contributed to the distribution of heat and resources in the crust. Furthermore, these answers will give us a better understanding of the role of fluids in initiating or influencing tectonic events. DISCUSSION: GEOLOGIC PHENOMENA ~ It is obvious from the above discussion that fluids within the crust are intimately involved in many processes of interest to geologists. For example, Hubbert and Rubey (1959) pointed out the mechanical problems associated with large overthrust sheets such as those observed in the Alps. They argue that one possible mechanical solution to large-scale overthrusting would be for associated pore fluids to have pressures approaching the lithostatic pressure exerted by the overlying rock. Under such conditions, the frictional resistance to sliding becomes negligible; overthrusting can occur as a gravitational process (sliding). While the original thought was to apply this idea to thrust sheets, the theory has been ex- tended to the general problem of frictional failure.

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 11 Indeed, the more we observe dynamic processes operating in the active tectonic areas of the Earth, the more we see additional evidence suggesting that the pore pressure in many, if not most, of these regions is high, well above hydrostatic. Unfortunately, most of the evi- dence is indirect, often only inferences from other geophysical observations. However, the current weight of the circumstantial evidence is such that a good case can be made for a rather general condition of high pore pressure in tectonically active areas. DRIVING FORCES The flow of fluid in response to a force is the central mechanism in the theory for single- phase nonisothermal, reactive transport in a porous medium. The potential driving forces provide considerable insight into the associated geologic phenomena. The force fields that drive flow are caused by the following mechanisms: topographic relief; tectonic dilation and compression; diagenesis; heat; and fluid source. Although there are other potential driving forces, such as chemical concentrations across clay membranes (osmosis) and electrical potential, they are small in relation to those listed above. The generation of pore pressure is a rate-controlled phenomenon; some mechanism, or set of mechanisms, operates to generate a hydraulic gradient. Pore pressure is dissipated by fluid flow outward from the source. The amount of the pore pressure increase depends on the ease with which flow can occur, which in turn depends on the hydraulic conductivity (permeability) of the host rock. As shown in Figure 2, observed permeabilities can range over 15 orders of magnitude within the crust. All of the major mechanisms that drive fluid flow (given above) have also been sug- gested to create high pore pressure in active tectonic regimes of the crust. Topographic Relief On the tectonically stable portions of the continents, most, if not all, groundwater flow is the result of topographic relief. Toth (1963) pointed out that the water table is the upper boundary for saturated groundwater flow, and that this boundary is usually closely approximated by the land surface. While the water table may fluctuate seasonally and from year to year because of wet and dry seasons, it does not fluctuate a great deal. Under most climatic regimes the water table can be demonstrated as approximately stationary over time. Forces associated with the groundwater table surface are important enough in under- standing groundwater flow that they should be stated another way. If we install a piezome- ter into the groundwater flow system within the Earth, we generally find the hydraulic head to be within 50 m of the land surface, certainly in most instances within 100 m. Hubbert and Rubey (1959) referred to this as the "normal" condition. This condition implies that the land surface is approximately the upper boundary condition for the groundwater flow system. The lower boundary for the system is no flow at some depth. The flow system is driven by topographic relief on the water table, the upper boundary for the saturated groundwater system. Lateral boundaries are formed by drainage divides that separate the flow system into discrete cells. Toth (1963) showed that the local topography provides perturbation on the regional flow system (Figure 3~. Because of significant scale effects with small flow

OCR for page 1
16 FIGURE 6 Schematic geologic cross section of the Skaergaard intrusion, showing the large-scale meteoric-hydrotherTnal circu- lation pattern in the plateau lavas above the granite gneiss base- ment (from Norton and Taylor, 1979~.. OVERVIEW AND RECOMMENDATIONS NO VERTICAL EXAGGERATION N 1~ O I ~ Kilometers -;: ~ rate of heat dissipation by circulating groundwater. In the lower and middle crust, the amount of groundwater that can be convected is poorly known. Fossil geothermal systems constitute several classes of hydrothermal ore deposits. In these systems, ore has been emplaced by the moving fluids. It seems possible that many vein-type ore deposits are formed in hydraulic fractures held open by the pore pressure during mineral deposition. Most sills and dikes seem also to be simple hydraulic fractures into which magma was injected. Geothermal power development exploits a concentrated heat source in the crust and depends on flowing groundwater (or steam). Whether or not commercial development of such a heat source is feasible depends almost entirely on whether the permeability of the hot rock is high enough to allow enough groundwater flow to replenish the heat efficiently. Geothermal reservoir engineering today involves numerical simulation of the reservoir's performance, which requires that the coupled partial differential equations be solved for the boundary conditions of interest. Source of Fluids What are the sources of fluid in the crust? Although fluids within a few thousand meters of the Earth's surface are likely to be derived primarily from precipitation, the source for fluids at deeper crustal levels is problematic. Fluids released from crystallizing magmas must contribute some fluid to the crustal fluid reservoir. Rocks that are heated as they are buried to deeper levels, or that are in the vicinity of intruding magma bodies, will also contribute fluids as they recrystallize to new mineral assemblages that are more compatible with their new thermal environment. (See Walther, Chapter 4, for additional discussion.) Finally, fluids may be released directly from the mantle to the overlying crust during "degassing" events. The relative proportions of these sources in any particular region have not been well established, although some isotopic studies have been undertaken to evaluate the relative roles of these processes. (See Taylor, Chapter 5, and Bredehoeft and Ingebrit- sen, Chapter 11, for additional discussion.) Hydration or dehydration of minerals changes the fluid mass in the pores and will change the pore pressure. Depending on the volume change associated with the dehydra- tion, such a source of fluids can increase the pore pressure. Perhaps the most commonly

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 17 discussed example is the montmorillonite-illite transformation that Burst (1969) first sug- gested as the explanation for high pore pressure in the Gulf Coast. Various higher- temperature metamorphic reactions release water. Most of these dehydration reactions are endothermic (i.e., absorbing heat), which slightly reduces the potential pore pressure buildup from the release of water. The magnitude of the fluid-pressure change that accompanies such a reaction is rate de- pendent. If fluid is generated at such a rate that it cannot flow readily away from the source, then fluid pressure will build up. If the permeability of the surrounding host rock is sufficiently small, the build-up in fluid pressure may be quite large. One other potential source of fluid flow into the crust is mantle degassing. Rubey (1951) presented the case that the volatiles associated with the Earth's surface were not present early in the life of the planet and have accumulated with time. Perhaps the most commonly accepted hypothesis is that the volatiles have been degassed from the mantle, Figure 7. The role of mantle degassing as a source of crustal fluids is reviewed in Chapter 11 by Bredehoeft and Ingebritsen. RECOMMENDED RESEARCH There are numerous important research topics involving the role of fluids in the crust. The list below is not all inclusive; it reflects those topics that we felt deserved special con- sideration. Each item is discussed in more detail following the list. 1. There should be continued studies of field areas where the role of fluids in geologic processes can be well documented, especially rock assemblages from deep in the crust. Of particular interest are the hot active areas the geothermal areas and the mid-ocean ridges. 2. Research is needed on the continued quantification of fluid transport in fractured rocks. Such research is important in understanding both crustal fluid flow and contaminant transport in shallower groundwater systems. 3. There needs to be a greater understanding of the nature of permeability, for example, the processes of diagenesis in altering permeability and the relationship of permeability to dispersion or mixing. 4. A greater understanding is needed of fluid flow through rocks of low permeability, for delineating the history of flow through sedimentary basins and for questions of the disposal of hazardous materials in the shallow crust. 5. Continued research is necessary on the kinetics of mineralogic reactions in the presence of fluid transport. Such reactions are important in understanding the rate of fluid phase generation. Not only should the classic mineralogical reactions be considered, but a greater emphasis should be placed on understanding the kinetics of organic reactions involving microbes with the mineralogical matrix. 6. There needs to be greater study of pore fluids in tectonically disturbed geologic re- gimes, both fossil and active regimes, particularly in situ sampling. 7. Research on the role of fluids in crustal deformation should be encouraged to help understand the time-dependent behavior of deformation and the role of fluids in rock me- chanics. 8. Advances are required in the mathematical simulation of coupled flow problems, which are now addressable using very large scale computational facilities and could supply an increased understanding of the physical processes and their prediction. 1. Field Studies Much of the theory concerning mass and energy transport by moving fluids is derived from laboratory and near surface investigations. The application of these ideas to the deep crust is speculative. New investigations often provide additional insights.

OCR for page 1
18 OVERVIEW AND RECOMMENDATIONS MANTLE DEGASSING -I ~ ~ .~; ~ 5 MAGMA DEGASSING '!Io Pilots ,- MELT EXTRACTION IN SITU GENERATION FIGURE 7 Schematic illustration of the possible sources for deep crustal fluids. ~ ~: t? , _ The application of the theory to real field areas is difficult and laborious. As Table 3 in- dicates, a large set of data must be assembled and evaluated quantitatively if one is to apply the theory rigorously. Commonly there is a lack of data to understand a wide range of fluid flow problems; some critical parameters must be evaluated by analogy to other areas where they are known. Often inverse methods are applied to determine the range of critical parameters. Because of the demands of data few attempts to apply mass and energy transport theory mathematically to analyze real field problems have been attempted. Even the attempts to simulate basin-scale fluid movements are limited in number. In dealing with the deep crust, knowledge of the fluids will be largely indirect. Our knowledge of the fluids will depend on our understanding of the petrology of the rocks and the role the fluids must have played. A number of studies of exposed deep metamorphic rocks indicate large fluxes of fluids; often these studies indicate that fluids several times the "pore volumes" circulated through the rock. Often the only remaining clues are isotopic signatures and sometimes fluid inclusions in the rock. It is from this scant indirect evidence that the properties and budgets of fluids and their effects must be inferred.

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 19 In the active areas one can study the processes operating currently. It is for this reason that the active geothermal areas are interesting. In these areas one can observe heat and mass transport properties directly along with concurrent chemical reactions. Much work has gone into modeling heat and fluid (including steam) movement in active geothermal areas. These reservoir simulations were partly initiated to evaluate power development; however, the principles are being applied to study geologic processes as well (see, for example, Bredehoeft and Ingebritsen, Chapter 11~. To date the geochemistry of the geothermal systems has not been integrated formally into an irreversible thermodynamic construct (see Norton, Chapter 2~. Ore deposits are often fossil geothermal systems. The discovery of "black smokers" at the mid-ocean ridge with 350C sulphide-rich water is evidence for large-scale geothermal circulation at active spreading centers through young oceanic crust. This hydrothermal circulation is almost certainly driven by high tem- peratures below the ridge. Evidence suggests that the deaths of fluid circulation mav have _ ~ ~ , , . , ~ . Em. . ~ . ~ . . _ been l to ~ km. '['here also may be mantle-derived volatile fluids as indicated by 3He contents of the fluids coming up along the mid-ocean ridges. The more actual field studies in which the fluids and their role are recognized and -__1_ __ I .1~ _ rat ~ . .~ .. . . . . . _ _. ~ _ analyzed, the more confident earth scientists can be in their understanding of the role of fluids. The application of the theory to the Earth is the experimental test. Often a careful field investigation provides new insight. 2. Fractured Rocks Numerous studies have indicated that the laboratory measurements of permeability do not characterize the flow properties of rock in situ. Often, in situ measurements indicated that the presence of fractures dominates the flow regime at a field scale. These fractures have a spacing larger than those in typical laboratory specimens. This is especially true for older well lithified sedimentary rocks as well as crystalline rocks. How to handle the problems of fluid migration and transport in a fractured medium is still an open question for both hydrologists and reservoir engineers. Two schools of approach have developed. The first school holds that the problem is tractable, provided one can describe the geometry of the fractures and the fracture network (see Figure 8~. This requires that the orientation, extent, and aperture of each fracture be described. Clearly, the problem of describing the fracture geometry is formidable, especially when this must be done at depth in the subsurface. Mathematically treating the fractures as frac- tals holds promise for greater characterization. The second school views the fractured rock as essentially a porous medium in which the pores are more widely spaced. The second group tends to describe the continuum in more classical terms as an anisotropic porous medium. Other characterizations include the so-called dual porosity medium in which flow in the interior blocks is characterized as flow in a porous medium while flow in the fractures is characterized by a much higher permeability, presumably representing the network of fractures. No consensus currently exists within the scientific community on how to deal with problems of fracture flow. Much more field laboratory and theoretical research must be done before our understanding will permit us to confidently treat this important problem. Much recent research has gone into studying fractures and attempting to characterize their distributions. 3. Permeability The rates of fluid migration depend on the permeability of the rock. Neglecting for the moment the role of fractures in permeability, discussed above, there are important research questions associated with the permeability of classical porous media. Permeability before diagenesis is controlled by the Ethology of the rocks. There are

OCR for page 1
20 OVERVIEW AND RECOMMENDATIONS it' \ ~ ?.. . ~1 ~ L/ : ~ Dl FIGURE 8 An example of a computer-generated model of a fracture network. Fractures are repre- sented as disc-shaped features of variable size. S(~ _y surprisingly few studies of the lithologic control on permeability. This is especially true when one looks at rocks of low permeability, the "confining layers" for the hydrologist, the "cap rock" of reservoir engineering. One major phenomenon that occurs in transport of chemical constituents in a porous me- dium is hydrodynamic dispersion, a process of physical mixing of the fluid through the matrix. This process has been shown to result from local small-scale variations in the fluid velocity field. Velocity variations in turn depend on the local variations in permeability. It is important, therefore, to understand the structure of permeability including its variabil- ity. Unfortunately, detailed data describing permeability, especially in situ permeability, is lacking. Research into the nature of permeability and its relationship to Ethology, includ- ing the diagenesis of rocks, is badly needed. Geostatistics is being applied both to the study of permeability and to dispersion; permeability distribution is also amenable to study by the fractal geometry. The application of these methods may provide new insights into how permeability varies. 4. Low-Permeability Rocks The role of low-permeability or "tight" rocks in controlling subsurface fluid movement is not well understood. Careful measurements need to be made in situ, but the measure- ment in rocks with such low permeability is difficult. For example, a borehole represents a potential short circuit for flow in the system, and this simple act of making the measure- ment might significantly disturb the system. Measurements in such a low permeability regime are at the level Heisenberg uncertainty. Bedded salt was thought by many to be impermeable because it is plastic and at some places it has existed since the Paleozoic era. Recent investigations at the Waste Isolation Pilot Plant (WIPP), which involve a mined nuclear repository in Permian Salado Formation salt, indicate that the salt is saturated with brine. One hypothesis for movement of the brine is that the salt behaves as any other porous material, except in this instance with very low permeability (10-8 to 10-9 Darcy). Shales also have very low permeability when measured in the laboratory. However, when the permeability is measured in situ where the scale of measurement is of the order of kilometers or larger, the permeability has been shown to be larger by two or more orders

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 21 of magnitude. This suggests that more conductive fractures exist, which may be spaced kilometers apart. Low-permeability formations isolate deeper formations from the surface and make possible lateral flow in deeper sedimentary basins. Without these tight formations fluids , ~ , _ would circulate down and back to the surface within short distances. Understanding and measuring the physical properties of the low permeability rocks is of great importance in gaining knowledge about regional subsurface fluid movement, which is critical for aD~li- cations such as the isolation of toxic wastes within the Earth's crust. 5. Chemical Reactions - -r r Moving fluids that transport both heat and mass facilitate the precipitation of minerals, dissolution of the rock matrix, ion exchange, and a host of potential chemical reactions. The total system is dynamic with reactions occurring in the moving fluids. The nature of the problem dictates that it is not enough to know which chemical reactions will occur (a problem in itself), but one must also understand the reaction kinetics if one is to make a quantitative analysis. Sometimes the flow is sufficiently large that assuming a local chemical equilibrium is adequate. In other instances, the rate of reaction is the controlling parameter. In the past the emphasis in geochemistry has been on understanding the inorganic com- pounds, the rock and mineral deposit forming minerals. It is increasingly clear that much of society's interest is in the organic compounds, especially hydrocarbons and a variety of contaminants. There are many unknowns about the role of organic geochemistry; even more poorly known are the catalytic properties of mineral surfaces and their effects on or- ganic reactions. In addition. some ore-formin~ minerals m~v he. tr~n~n~rt~rl ~c nrn~ni~ complexes in aqueous systems. Hydrologists are becoming increasingly aware that microbes exist in the subsurface en- vironment. Most of the chemical reactions associated with organic groundwater contami- nation are controlled biologically. The microbes appear to exist in the subsurface in a state of near starvation. Any potential food source is immediately seized upon. It appears that a rapid biologic adaption is possible to accommodate the potential food source. The microbes have been found in the groundwater-saturated subsurface to depths approaching a kilometer; but the maximum depth at which they can exist is not clear. The entire area of geochemistry, especially organic geochemistry, in a moving subsur- face fluid system is ripe for exciting new research. At shallow depths, organic reactions are clearly biologically controlled. It is not clear to what depths the biota play a significant role. ~ _ ~a ~A^Vr~^ ~ ~4 t>~AA4- 6. Pore Fluids in Active Tectonic Areas 1:~^ ~O cat 1~^ i_ ~1~^ ~ ,~1 +~ 1~ .. . ~ _ _ 1 ~ _ 1_ _ _ ~ , ~ Vim ~l~Ul~ ~ QU~1 V~U [U ~O VOly Lilly In a number OI pelrOleUm prOVlIlCeS. However, empirical observations in tectonically active areas where oil and gas are not known to be present are limited to very shallow depths, usually less than 300 m (approxi- mately 1000 feet). Indirect methods suggest that many of the tectonically active areas may have pore pressures approaching lithostatic. However, one needs empirical observations to corroborate geophysical and other indirect data. A program of deep drilling for scientific purposes would greatly add to our empirical knowledge concerning pore fluids, and, in turn, their role in active tectonic processes. The few drillholes that have sampled fluids at depth-e.g., at the Kola Peninsula in the USSR and at Cajon Pass in California have provided some intriguing data. However, in many areas of great scientific and societal interest there is no direct information at depth. Pore fluid may play an important role in failure of the Earth materials. The fluid state may be most important in understanding the mechanics of deformation including earthquakes.

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 7. Crustal Deformation Hubbert and Rubey (1959) indicated that fluids can play an important role in the faulting process. Subsequent investigations into controlling earthquakes at Rangely, Colorado, demonstrated the applicability of the Hubbert and Rubey hypothesis. This hypothesis has only been applied to a linear elastic theory; earth materials are known to be viscoelastic. Recent development of constitutive laws for rocks undergoing failure include viscous effects. What we know of the earthquake failure mechanism is best explained by a viscoelastic failure behavior. The role of pore fluids in the newer models for viscoelastic rock behavior is an area for continued research. The exact rock behavior prior to and during an earthquake is not well known. "Is there dilatancy; and what is its effect on failure?" is still the subject of debate and investigation. There may need to be observations of pore pressure within the focal zone during an earthquake to provide conclusive information a formidable task of both predicting an earthquake and making the measurements. This area of the role of fluids in the deformation and failure process of rocks is one of continuing investigation and interest. Results from the earthquake prediction experiment at Parkfield~ California, may provide new insight. 8. Mathematical Analysis Modeling of the roles of fluids in crustal processes and prediction of how fluids (espe- cially those containing contaminants) move is a broad goal for the earth scientists. The set of partial differential equations referred to above and described in detail in Chapter 1 incor- porate our fundamental understanding of the physics and chemistry associated with the role of fluids in the crust. Solutions to the basic equations provide insight into how the systems operate under a variety of complex geologic settings. Only in the simplest cases are analytical solutions to a problem of interest feasible. However, the digital computer has made it feasible to solve realistic problems of interest. As the computers become more powerful the problems addressed have also become more realistic. The numerical methods commonly in use e.g., finite difference, finite element, and method of characteristics- have become household words in the geosciences. Investigators now solve transient-flow problems in two- and three-space dimensions routinely, in both petroleum engineering and groundwater. There are, however, problems of interest that tax even the largest computers. The approach to mass and heat transport in the geosciences has adopted methods of nonequilib- rium thermodynamics developed for chemical engineering; this approach, outlined in Table 2, involves writing a separate equation for fluid pressure and temperature and for each chemical constituent of interest. A set of partial differential equations is then solved simultaneously. For the time-dependent problem in three-space dimensions, this becomes a very large computational problem. Such problems can involve thousands of CPU hours on even the largest available super computers. There is an obvious need in the earth sciences for very large scale computing facilities. A critical part of the analytical methods used in geohydrology is the comparison of field data to analytical results; in petroleum reservoir engineering this comparison is referred to as "history matching" perhaps its best description (other disciplines, such as hydrology, use other terms to describe this comparison such as calibration or parameter estimation, which give the comparison an aura of more reality than it deserves). The history match is all important for real problems. It addresses the question, can the system as the investigator thinks he understands it, produce this set of observations? Most real geologic systems are sufficiently unconstrained by information so that no unique solution is possible. One is usually led to ask the question, is there a set of reasonable parameters that describe the system and yield something close to the observed

OCR for page 1
OVERVIEW AND RECOMMENDATIONS 23 result? In actual fact, one is usually left to play with the analytical model of the system to see which parameters the system behavior is sensitive to, this procedure is often referred to as sensitivity analysis. Sensitivity analysis can be performed by trial and error, or some more rigorous procedure. In the end, the investigator gains a feeling for just how the system responds. Since the information is incomplete with respect to uniqueness, one is often left to make a professional judgment as to the actual nature of (1) fundamental rela- tionships that describe the system, (2) boundary and initial conditions, and (3) the parame- ter distribution within the system. The parameter distribution of most geologic problems is described by a very limited sample. For problems at depth within the Earth, drilling and sampling, while expensive, are a critical endeavor to permit the state of parameter estimation and history matching. Since the system behavior depends on the parameter distribution within the system, one would like to look at the effect upon the system of variations in parameter distribution. Past investigations have generally tested what was thought to be the best single interpretation of the parameters of interest; for example, the permeability distribution. Given this distnbu- tion, the system behavior was computed. However, one would like to place a confidence band about the computed system behavior. Given confidence bands about system re- sponse, one is in a better position to judge the "goodness" of the history match. However, in the end the "goodness" of the history match is a judgment. One advantage of using a time-dependent mathematical analysis is that one can make a quantitative prediction of system behavior. This is particularly useful for environmental problems. For example, one would like to know how a plume of contaminated groundwa- ter is moving in an aquifer, in order to take action. Future research in the geosciences must be directed toward placing confidence bands about the output of analytical models, including predictions. REFERENCES Brace, W. F. (1980~. Permeability of crystalline and argillaceous rocks: Status and problems, Inter national Journal of Rock Mechanics in Mineral Sciences and Geomechanical Abstracts 17, 876- 893. Bredehoeft, J. D., and B. B. Hanshaw (19681. On the maintenance of anomalous fluid pressure, I. Thick sedimentary sequences, Geological Society of America Bulletin 79, 1097-1106. Bredehoeft, I. D., C. E. Neuzil, and P. C. D. Milly (1983~. Regional flow in the Dakota aquifer, U.S. Geol. Survey Water Supply Paper 2237, pp. 1-45. Burst, J. F. (1969~. Diagenesis of Gulf Coast clayey sediments and its possible relation to petroleum migration, American Association of Petroleum Geologists Bulletin 53, 73-79. Fyfe, W. S., N. J. Price, and A. B. Thompson (1978~. Fluids in the Earth's Crust, Elsevier, Amster dam, 383 pp. Hubbert, M. K., and W. W. Rubey (19591. Role of fluid pressure in mechanics of overthrust faulting, Geological Society of America Bulletin 70, 1 15-166. Norton, D. (1984~. A theory of hydrothermal systems, Annual Reviews of Earth and Planetary Sciences 12, 155-177. Norton, D., and H. P. Taylor, Jr. (1979~. Quantitative simulation of the hydrothermal systems of crystallizing magmas on the basis of transport theory and oxygen isotope data: An analysis of the Skaergaard intrusion, Journal of Petrology 20, 421-486. Norton, D., H. P. Taylor, Jr., and D. K. Bird (19841. The geometry and high-temperature brittle deformation of the Skaergaard intrusion, Journal of Geophysical Research 89, 10,178-10,192. Palciauskas, V. V., and P. A. Domenico (1982~. Characterization of drained and undrained response of thermally loaded repository rocks, Water Resources Research 18, 281-290. Rubey, W. W. (195 1~. Geologic history of the sea, Geological Society of America Bulletin 87, 1111- 1148. Toth, J. (1963~. A theoretical analysis of ground-water flow in small drainage basins, Journal of Geophysical Research 68, 4795-4812.

OCR for page 1

OCR for page 1
BACKGROUND

OCR for page 1