Subproject G5: Dehydration and partial melting as consequences of mass transport in the deep Andes - constraints from petrophysics
Research Field: Geophysics, petrophysics, laboratory studies, fluids/melts
-
Research staff
-
Working area
-
Abstract
-
Objectives, methods, work plan, and schedule
-
Collaboration with external research groups
-
Publications
Staff
Project-Leader(s)
Co-Leader(s)
Working area of the subproject G5
Fig. 1: Working area of project G5
Abstract
The central question of the proposed project is how petrophysical properties of rocks typically occurring in the Central Andean crust change with dehydration and/or partial melting reactions: how the liquids affect physical properties and how they finally escape from the reacting system. As fluids play an important role in the current explanations of geophysical observations in the Central Andes, it is imperative to have high quality laboratory data to constrain these observations. In the current phase of the SFB, we investigate the petrophysical properties of dehydrating serpentinite at "in situ" conditions. For the next phase we are planning to extend the scope of the project to more complex metamorphic crustal rocks of felsic, intermediate or mafic compositions. Water released during the dehydration reaction triggers partial melting in the lower and middle crust. For this reason, dehydration and the transport of water as well as metamorphic reactions initiated by the influx of water are key parameters for many processes in general and in the Andes in particular. Electrical conductivity is ideally suited to measure the effects and the distribution of the released water; therefore the emphasis of this study will be on the determination of electrical conductivity of dehydrating and/or partially molten rocks at high pressure and temperature. The dehydration induced microstructure and the pathways for fluids evolved after the reaction will be analysed in detail after the experiments by optical microscopy as well as by scanning and transmission electron microscopy (SEM and TEM). With the application of appropriate model assumptions we will then be able to estimate the range of fluid transport in time and space. These estimates will be combined with geophysical field measurements (G1, G2, G3, G4). Elastic velocities (Vp, Vs) and thermal properties (conductivity, diffusivity) will be determined for the same sample materials.
Objectives, methods, work plan, and schedule
It is one of the major goals of SFB 267 to synthesise data as well as to model and simulate related processes in order to identify key parameters and their interaction for various styles of subduction orogeny at an Andean-type margin. As the influence of fluids and/or melts in weakening crustal materials has been identified as one of the key parameters in controlling subduction orogeny, the goal of the proposed investigation is to determine the petrophysical properties of rocks of the Andean middle crust during ongoing metamorphic dehydration and melting reactions. Fluids and/or melts play an important role in the current explanations of geophysical observations in the Central Andes, so it is imperative to have high quality laboratory data to constrain those observations. One of the key parameters is to understand and quantify how the fluids and/or melts affect petrophysical properties at the various stages of the reaction and how they develop a connectivity in and finally escape from the reacting system. For this purpose primarily electrical conductivity but also seismic velocities and thermal conductivity will be measured during the ongoing reaction.
Geophysical field observations can often be interpreted by more than one explanation. For example, high electrical conductivities can be caused by the presence of silicate melts, graphite or ore mineralisations, or hydrous fluids. The presence of abundant ore minerals and/or graphite would conflict with other geophysical and petrological observations, such as the seismic velocities or the observed mineral parageneses. But the seismic and magnetotelluric anomalies observed in forearc and Western Cordillera could be explained by either partial melting or a high fluid content.
For the quantitative interpretation of the processes occurring in the Andean crust and of the geophysical field observations, the available petrophysical parameters are not satisfactory. New experiments are therefore of major importance and highly relevant for the overall objective of the Sonderforschungsbereich. In addition to the variables temperature, pressure, and composition, the fluid flow is a parameter that needs to be determined for the analysis of rock properties. Because of the specific role of fluid- and melt-enhanced processes in the Andes, the determination of electrical conductivity is particularly important for the localisation, quantification and interpretation of ongoing processes in a subduction orogen. Electrical properties of rocks respond highly sensitively to transiently available fluids, which are also detectable over very short time spans. For this reason, we propose to measure the effect of different parameters (pressure, temperature, connectivity and composition of fluids in the pore volume) on the electrical conductivity of dehydrating rocks, and its evolution with time in laboratory experiments. The laboratory experiments proposed here can contribute significantly to the clarification of the source of the observed anomalies. We will be able to quantify how fluids and/or melts affect the petrophysical properties and how the physical properties change with ongoing metamorphic reactions. These measurements can then be used to evaluate the geophysical field observations. Electrical conductivity measurements can be used to quantify such parameters as connectivity of fluids or melts.
The connectivity of free hydrous fluids in rocks is quite different from that of silicate melts in partially molten rocks. The degree of connectivity of the liquid phase depends critically on its wetting behaviour, which is a function of the interfacial energies of the solid/liquid system. The rate of extraction of a liquid phase at the meso- and macroscale strongly depends on the connectivity at the microscale as well as on density and viscosity of the liquid phase. Since buoyancy forces are the driving force for the ascent of liquids at the macroscale, liquid segregation rates depend strongly on the density contrast between the liquid and the solid phase. In addition, silica rich melts have a viscosity that is several orders of magnitude higher than that of hydrous fluids, and thus mobility of melts will be much lower than that of fluids. It is therefore expected that fluids and melts will have distinctly different effects on the electrical conductivity of a rock at all scales.
Information gained from the experiments proposed here will answer some basic questions on fluid flow in lower and middle crustal rocks and will be used in models addressing mass transport and the evolution of the Andean orogen (projects F1, G1, G2). Thus these measurements will greatly improve our understanding of the mechanisms of fluid and melt generation and flow in the lower and middle crust in general and of the Andes in particular.
Methods
Measurements of the electrical conductivity will be performed in an internally heated gas apparatus especially designed at GFZ for rock physics experiments using large volume samples. The apparatus is used to determine electrical conductivities of serpentinite rocks in the current period of the Sonderforschungsbereich. The maximum confining pressure is 1.4 GPa and the maximum pressure of the second pressure system controlling the fluid (pore) pressure is 0.7 GPa, such that conductivities may be measured at controlled pore pressures. The maximum temperature is 1000?C. Seismic velocities and thermal properties can be measured on the same materials in the same apparatus.
Two other pressure vessels are available for large volume samples, one for experiments to determine elastic properties at pressures up to 1 GPa and temperatures up to 1600?C, and the other for highest precision electrical conductivity measurements up to 250?C and 300 MPa confining pressure with simultaneous control of pore pressure.
Electrical conductivity measurements will be performed with a new impedance spectrometer that allows measurements even at the very high resistivities of cold rocks but also for the low resistivities expected for water-saturated hot rocks. The frequencies used for conductivity measurements will range from 10 mHz to1 MHz. This wide range of frequencies allows the separation of grain boundary and bulk conductivities, which is a basic requirement for the understanding of the underlying processes leading to the observed conductivity as well as for the scaling of laboratory measurements to field observations in the Andes.
The rocks used for these experiments will be representative of the Andean lower and middle crust, such as amphibolites or micaceous metagraywackes (biotite-plagioclase gneisses). Sample cylinders will be cored from natural rocks and then cut and polished to exact sample dimensions (e.g., 75 mm long, 29 mm diameter). The dehydration induced microstructures and fluid and/or melt connectivity will be studied by detailed optical microscopy, SEM and HRTEM (high resolution TEM) and EELS (electron energy loss spectroscopy). Using appropriate model assumptions, it will then be possible to estimate the reach of the fluid and/or melt transport in space and time.
For our first experiments we plan to use an amphibolite (Western desert Egypt, so-called Chephren-Amphibolite) petrologically characterized by Franz (1987). Amphibolite is a representative material because it is an important constituent of the lower crust in the Andes and probably a major constituent in the middle crust of the forearc. The Chephren amphibolite is an ideal starting material as it is composed of mainly two phases, amphibole and plagioclase, and is well characterised and only minimally affected by retrograde greenschist metamorphism. For amphibolites, the phase transition of amphibole (hornblende) to clinopyroxene and plagioclase begins at approximately 775?C on the quartz-fayalite-magnetite buffer (Spear, 1993); the upper terminal stability of hornblende is over 900?C, where the transition to clinopyroxene + orthopyroxene + plagioclase + olivine occurs. The solidus of amphibolites is strongly pressure dependent, with temperatures above 900?C for very low pressures (< 200 MPa), and temperatures below 750?C for pressures of 1 GPa (Wolf & Wyllie, 1993, 1994). Such a low solidus implies that dehydration of hornblende will lead to dehydration melting at these higher pressures. Thus amphibolite is an ideal sample material for the proposed study, not only because it is a rock representative for the lower and middle crust of the Andes, but also because two different scenarios can be investigated with the same sample material: the effect of water produced by dehydration and the effect of melts.
Depending on the success of the experiments, we are considering the investigation of a second representative rock type, biotite-plagioclase gneiss, which dehydrates at lower temperatures and is more representative for the middle to upper crust. Such already well characterised material is also available from the group G1. The results of this study will therefore be of major importance for the interpretation of geophysical field observations where the effects of water and melt need to be distinguished.
Experimental strategy
As the exact dehydration and melting conditions of a material depend on its composition, a first petrological experiment is planned to verify the P,T,t-conditions for the onset and the maximum of the dehydration reaction of the sample materials. For electrical conductivity measurements on amphibolites, we intend to investigate first the effects of hornblende dehydration, which requires low pressures if dehydration melting is to be avoided. We plan to raise the pressure first to 300 MPa, in order to suppress thermal cracking during the subsequent heating stage. Temperature will be raised in two stages. In stage 1, the temperature will be raised at a rate of 1?C/min. up to 700?C. Temperature will be raised at a slower rate above 700?C up to 900?C, where the overstepping of the equilibrium temperature grants dehydration kinetics that allow observation of the reaction at laboratory timescales (i.e., up to 48 hours), but where no melting occurs. Sample resistivity will be monitored continuously with an automated impedance spectrometer. Measurement of one impedance spectrum takes one to two minutes, depending on the frequency range.
In a second experiment, pressure will be raised to 1 GPa to investigate the effect of dehydration melting on the electrical conductivity at similar temperature ranges as for the first experiment. In subsequent experiments the change of resistivity with time will be monitored at varying P,T conditions for several hours, until no further change can be observed. On the basis of the results achieved in our working groups for invariant systems such as the serpentine to forsterite + H2O reaction, we expect the resistivity to change dramatically with dehydration. To investigate the effect of melts, similar experiments will be performed at pressures of 1 GPa. For strongly anisotropic materials, sample cylinders will be cored parallel and perpendicular to the foliation or the preferred orientation of the minerals.
Before and after the experiments, sample composition and microstructures will be analysed in detail to determine the effects of microstructural anisotropies and composition on the conductivity of the sample and to correlate the microstructure with the measurements. The interconnectivity of the pore volume will be microscopically quantified and reconstructed in three dimensions in a procedure similar to that described by Bruhn et al. (2000). Fracture networks can be evaluated and their influence on the permeability evolution can be estimated by the approach used by Huenges & Zimmermann (1999), who modelled the extrapolation of microscale permeabilities determined in the laboratory to mesoscale permeabilities observed in the drill hole. Extrapolation to crustal scales will be performed in conjunction with field measurements obtained in project G4 (Brasse/Haak).
A PhD-student position requested for project G4 is intended to provide the link between the two projects in order to develop a model approach that will allow the eventual extrapolation of conductivities determined in the laboratory to the kilometre scale observed in the Andes. When the dependence of the connectivity of the fluids and/or melts on wetting behaviour at the microscale and physical parameters such as density and viscosity are appropriately considered, it will then be possible to characterise the nature, volume and distribution of the fluids or melts that are assumed to cause the anomalous conductivities measured by magnetotelluric methods in the Central Andes.
References
Bruhn, D., Groebner, N., and Kohlstedt, D.L. (2000) An interconnected network of core-forming melts produced by shear deformation. Nature, vol.403, 883-886.
Franz, G. (1987) Breakdown of amphibole and the formation of Al-titanite - an example from the polymetamorphic "Chephren Diorite" (Gebl el Asr, SW Egypt). 14th Coll. African Geology, Techn. Univ. Berlin. Publication Occ., CIFEG Paris, Abstr. Vol p 76
Huenges, E. & Zimmermann, G. (1999) Rock permeability and fluid pressure at the KTB. Implications from laboratory - and drill hole - measurements. Oil & Gas Science and Technology - Rev.IFP, vol. 54, No.6, 689-694.
Spear (1993) Metamorphic phase equilibria and pressure-temperature-time paths. Mineralogical Society of America, Washington. 799p.
Wolf, M.B. & Wyllie, P.J. (1993) Amphibolite dehydration-melting: sorting out the solidus. in Prichard, H.M., Alabaster, T., Harris, N.B.W. & Neary, C.R. (eds.) Magmatic Processes and Plate Tectonics, Geol.Soc.Spec.Publ. No. 76, 405-416.
Wolf, M.B. & Wyllie, P.J. (1994) Dehydration melting of amphibolite at 10 kbar: the effects of temperature and time. Contrib. Minral. Petrol., 155, 369-383.
Work plan and schedule
Within the first year, composition and microstructures of the potential starting materials will be completely characterised. Petrologic experiments to determine the kinetics of the reactions will be performed to fix the experimental conditions for the petrophysical experiments.
Petrophysical experiments will be performed in the first and until the end of the second year. First experiments will be under undrained conditions, such that all the fluids stay in the system. Later on, the pore pressures will be altered to determine the dependence of the observed effects on pore pressure. Each experiment will be thoroughly analysed and interpreted, to make sure results are available for numerical models. Preparation of these highly complex experiments takes four to six weeks. Thus we are planning to perform eight to twelve experiments in each the first and the second year. As laid out in Experimental Strategy above, the first two experiments will determine the range of sample resistivities at the conditions of the reaction. The exact procedure and design of the following experiments will therefore be decided after these two experiments. Experimental setup needs to be modified for each experiment to optimise the design for each material and experimental conditions. The next experiments will address the evolution of petrophysical properties with time at ongoing dehydration and/or melting reactions.
The last year will be used for postexperimental investigation and interpretation and synthesis of the data in the context of geophysical field observations. Thus data will be interpreted together with projects G4 (Brasse/Haak), G1 (Franz/Sobolev) and G2 (Handy et al.).
Collaboration with external research groups
Publications
Literature
Reviewed publications
Arndt J., Bartel T., Scheuber E., Schilling F.R. (1997): Thermal and Rheological Properties of Granodioritic Rocks from the Central Andes, North Chile. - Tectonophysics, 271: 75-88; . - []
Giese P., Scheuber E., Schilling F.R., Schmitz M., Wigger P. (1999): Crustal Thickening Processes in the Central Andes and the Different Natures of the Moho-Discontinuity. - Journal of South American Earth Sciences, 12: 201-220; . - []
Schilling F.R., Hauser M., Sinogeikin S.V., Bass J.D. (2001): Compositional Dependence of Elastic Properties and Density of Glasses in the System Anorthite-Diopside-Forsterite. - Contrib. Mineral. Petrol, 141: 297-306; . - []
Schilling F.R., Partzsch G.M. (2001): Quantifying Partial Melt Portion in the Crust Beneath the Central Andes and the Tibetan Plateau. - Physics and Chemistry of the Earth (A), 26: 239-246; . - []
Schilling F.R., Sinogeikin S.V., Bass J.D. (2003): Single-Crystal Elastic Properties of Lawsonite and their Variation with Temperature. - Physics of the Earth and Planetary Interiors, : ; . - []
Schilling F.R., Sinogeikin S.V., Hauser M., Bass J.D. (2003): The Elastic Properties of Modal Basaltic Melt-Compositions at High Temperatures. - Journal of Geophysical Research.
Seipold U., Schilling F.R. (2003): Heat Transport in Serpentinite. - Tectonophysics, : ; . - []
Sinogeikin S., Schilling F.R., Bass J.D. (2000): Single Crystal Elasticity of Lawsonite. - American Mineralogist, 85: 1834-1837; . - []
Sinogeikin S., Schilling F.R., Bass J.D. (2000): Single Crystal Elasticity of Lawsonite. - American Mineralogist, 85: 1834-1837; . - []
Yang X., Jin Z., Huenges E., Schilling F.R., Wunder B. (2001): Experimental Studies on Dehydration Melting of Biotite-Plagioclase Gneiss Under Granulite Facies Conditions. . - Chinese Science Bulletin, 46: 867-871; . - []
Qualifikationsarbeiten (Habilitationsschriften, Promotionen, Diplomarbeiten) und interne Berichte
Schilling F.R. (1998): Petrophysik ? ein mineralogischer Ansatz. - Habilitationschrift; . - []