ELECTRICAL CONDUCTIVITY MEASUREMENT OF GRANULITE SAMPLE UNDER LOWER CRUSTAL PRESSURE- TEMPERATURE CONDITIONS

Transcription

ELECTRICAL CONDUCTIVITY MEASUREMENT OF GRANULITE SAMPLE UNDER LOWER CRUSTAL PRESSURE- TEMPERATURE CONDITIONS
Session EM1: Laboratory Measurements
EM1-1
ELECTRICAL CONDUCTIVITY MEASUREMENT OF
GRANULITE SAMPLE UNDER LOWER CRUSTAL PRESSURETEMPERATURE CONDITIONS
Kiyoshi Fuji-ta (Kobe University, Department of Earth and Planetary Sciences,
1-1 Rokkodai, Nada, Kobe, Japan, e-mail:fuji-ta@kobe-u.ac.jp)
Tomoo Katsura (ISEI, Okayama University, Japan)
Yoshiaki Tainosho (Faculty of Human Development , Kobe University, Japan)
A number of discussions have been done on the cause of Highly Conductive Layer
(HCL) in the lower crust. Most of laboratory experiments simulated the rocks in the
stable continent saturated by saline water, water or conductive minerals to account for
the HCL. On the other hand, the laboratory measurements of relatively dry rocks or
homogeneous rocks are not sufficiently done nor properly evaluated. We attempted,
therefore, to establish the technique to measure the electrical conductivity of real crustal
rocks with relatively low conductivity and complicated mineral components in order to
compare with the results given by the Magneto-Telluric (MT) measurements. A
granulite sample, representing the lower crustal rock, was obtained from Hidaka
Metamorphic Belt (HMB) in Hokkaido, Japan. The granulite sample was sintered
under the conditions similar to those of lower crust, and after achieving characterization,
the electrical conductivity of synthesized granulite sample was measured. The stable,
reversible and reproducible conductivity values were successfully obtained up to about
1000 K under the fixed pressure of 1 GPa. The result accorded with the electrical
conductivity structure suggested by the MT data analysis. In conclusion, to consider
pore fluid conduction mechanism or the role of accessory minerals in the rock, the
mechanisms of electrical conductivity paths in dry or basic rocks should be
reconsidered again.
EM1-2
LABORATORY MEASUREMENTS OF FREQUENCY DISPERSION OF
SEDIMENTARY ROCKS ELECTRIC PROPERTIES AS APPLIED TO GROUND
WATER HYDROCARBON CONTAMINATION
Hallbauer-Zadorozhnaya V., (Institute for Polimer Science of Stellenbosch University, South
Africa, dkh@mweb.co.za), Kamenetsky F. and Shmidbauer E. (Institute of Geophysics of
Munich Ludwig-Maximillian University)
Preliminary data of laboratory measurements of sedimentary rocks complex admittivity
γ (ω ) = σ (ω ) − iωε (ω ) obtained in Institute of Geophysics of Munich University in MarchApril 2002 are reported. The equipment used has been previously developed by Prof. E.
Schmidbauer for investigating the complex resistivity of minerals on the basis of the HewlettPackard high-precision LCR-meter HP4284A with frequency range 20 Hz to 1 MHz. The
method of variation of samples length has been used to eliminate the influence of electrode
polarisation. Several water saturated samples of sandstone and limestone have been measured.
The unconsolidated sample of Bavarian sand has been also investigated. It is known from
previous field and laboratory experiments that groundwater contamination due to hydrocarbon
compounds can essentially change conductivity of rocks and its dependency on frequency.
That is way some of the same samples were measured after partial saturation with Diesel fuel.
At the initial stage of research all the necessary qualitative co-relations could be hardly
observed. Nevertheless, a quite certain quantitative conclusions are obtained. Water saturated
consolidated and unconsolidated samples display no or very low conductivity dispersion.
Contaminated samples of unconsolidated sand display no dispersion. Contaminated samples
of limestone (porosity about 15%) display high conductivity dispersion: about 10.3 times
increase of conductivity with frequency. Contaminated samples of sandstone (porosity about
7%) display very high conductivity dispersion: about 320 times increase of conductivity with
frequency. Preliminary results obtained promise a new possibility to locate contaminated
areas by investigation conductivity dispersion or IP effects in geo-electromagnetics.
Frequency dispersion
of conductivity of
sedimentary rocks
samples
14
12
Conductivity mS
10
Contaminated:
1 — limestone σ( ⋅50 is
shown),
2 — sandstone (σ100
⋅ is
shown),
3 — sand with 20%
water and 10% Diesel.
8
6
4
2
0
1,E+01
1,E+02
1,E+03
1,E+04
1,E+05
Frequency, Hz
1
2
3
4
5
6
1,E+06
Water saturated:
4 — limestone,
5 — sandstone,
6 - sand with 30%
water.
EM1-3
TEMPERATURE, MINERAL COMPOSITION AND ELECTRICAL CONDUCTIVITY OF THE MANTLE
Juanjo Ledo and Alan G. Jones. (Geological Survey of Canada, 615 Booth St, Ottawa K1A 0E9, ON,
Canada). ledo@cg.nrcan.gc.ca
Using laboratory-derived temperature-dependence of electrical conductivity of mantle
minerals, we present physical and statistical methods to determine bulk conductivity of mantle
mineral assemblages for the ternary Olivine-Orthopyroxene-Clinopyroxene (Ol-Opx-Cpx) system. We determine the physical properties of composite materials by combining variational
bounds and percolation theory. We calculate physical properties bounds as a function of the fraction of different phases present; these limits correspond to the extreme situations where the most
conducting phase is either interconnected or disconnected. Percolation theory is used to determine
the probability of having a connected or interconnected conducting phase depending on the
dominant lattice geometry. The bulk physical property is then calculated through a weighted
arithmetic mean of the physical bounds. The relationships presented here between temperature,
mineral composition and electrical conductivity allows constraining one of them given the other
two.
As an example of the application of our approach, we discuss an area in the Yukon,
northwestern Canada, where there is xenolith evidence for a bimodal upper mantle mineral assemblage (harzburgite and lherzolite). This locality coincides spatially with an upper mantle region of low Vp, determined by teleseismic data, and low electrical conductivity, determined by
long period magnetotelluric data. The teleseismic data were taken to suggest an up to 200 °C localized increase in mantle temperature, compared to neighbouring regions that are more lherzolitic (Cpx-rich) in nature. Our mixing results can exclude this interpretation, suggesting that compositional variation would be a more appropriated explanation.
EM1-4
SCALING OF APPARENT RESISTIVITY OF FERROUS QUARTZITES
S.S.Krylov, (St.Petersburg State University. St.Petersburg, Russia, krylov@geo.phys.spbu.ru)
N.Yu.Bobrov, V.A.Lubchich (St.Petersburg State University. St.Petersburg, Russia)
Ln < ρa >
As it was demonstrated by the set of investigations, the banded iron formations (BIF) have
a specific fractal structure. To a first approximation these rocks can be considered as a bicomponent media composed of conductor (magnetite) and insulator (quartz). Such structures can
be studied using the percolation theory approaches. It is well known that the effective
conductivity depends on the system size according to the power low at the distances less than
correlation length of the system. We have performed the laboratory investigations of DC
resistivity of the sample of ferrous quartzite from the Pechguba BIF deposit (Kola Peninsula). The
sample had quasi-plane shape and its sizes were approximately equal to 0.9х0.7х0.2 м. The
apparent resistivity was measured at the upper polished plane with the Wenner’s array. This plane
was approximately perpendicular to the quartz bands. The electrodes were placed at the nodes of
rectangular grid with the cell size of 1 cm. The distance between the electrodes was varied from 1
to 10 cm. For each fixed distance about 100 measurements in the different points were carried out.
Apparent resistivity was derived from the measured values of voltage and current. The values
obtained at each distance were then averaged and log-log plot of resistivity mean value versus
distance was drawn. This dependence occured to be best fitted by the straight line. The correlation
coefficient of the least-square linear fit exceeded 0.99. The
18
power exponent µa was 1.49±0.04 for the four-electrode array
and 1.61±0.06 for the three-electrode array.
17
Then the electronic two-color image of the sample’s
surface was obtained with the use of a scanner. This image
16
was used for the calculation of fractal dimension by the boxµa=1.49
counting method and effective conductivity by the random15
R2=0.994
walks method. It is obvious that the results of these
calculations are to strongly depend on the brightness of the
image. In our work the brightness was selected in such a way,
14
that the calculated power exponents to coincide with the
0
0.4
0.8
1.2
1.6
2
Ln L
other ones determined in the experiment (taking into account
Logarithm of average apparent resisrivity
that not the conductivity but the apparent resistivity was
versus logarithm of distanse.
measured). For the selected image the fractal dimension
(four-electrode array)
evaluated by the box method is equal to 1.816, that is in a
good agreement with the value 1.83 obtained earlier by G.Yu.Ivanyk.
During the field work at the Pechguba BIF deposit TEM soundings with coinciding
horizontal loops have been carried out and anomalous polarization of ferrous quartzites have been
revealed. Analysis of the field sounding curves showed that the dependence of dielectric constant
on the frequency could be described by the “universal” Jonscher’s function. The assessment of the
crossover frequency for this model is compartible with the frequency, for which the value of
electromagnetic wave length is corresponding to the correlation length evaluated by the
measurements of apparent resistivity scaling.
Acknowledgements: The presented investigations were partly financed by the grant of
Russian Ministry of Education and Federal Program “Integratsiya”.
EM1-5
PROMOTION OF GRAPHITE FORMATION BY TECTONIC STRESS A LABORATORY EXPERIMENT
Georg Nover (Mineralogisches Institut, Universiitat Bonn , g.nover@uni-bonn.de)
Johannes Stoll (Institut fur Geophysik, Universittat Goettingen)
Graphitisation of less ordered hexagonal carbon was studied under in-situ pressure and
temperature conditions on anthracite, black shale and a synthetic calcite/anthracite mixture at
upper greenschist facies conditions. Anthracite exhibited a continuous loss of volatiles in the
temperature range from 100C up to 850C (9.9 weight % at 450C) as detected by
Differential-Thermo-Analysis (DTA) and Thermo-Gravimetry (TG). Energy dispersive Xray diffraction (EDX) revealed a broad amorphous 002 graphite reflection while after p,Ttreatment nearly perfect crystallised graphitic carbon was detected. The electrical
conductivity was measured at the same time in the frequency range from 0.7 up to 100 kHz.
As a function of time the bulk resistivity was decreased by about three orders in magnitude at
constant pressure and temperature conditions (0.7 GPa, 450C), while the complex response
exhibited a continuous decrease of the imaginary part of the impedance. "Quasi-metallic"
conduction now dominates the charge transport. Application of pressure, stress, temperature
and time caused an increase in ordering and the degree of interconnection of the formerly
random oriented carbon sheets. Thus graphitization requires temperatures in the order of 300500C and pressures of several 100 MPa. This result corresponds with the occurence of
graphite in overthrusts and nappe structures.
EM1-6
INFERENCES OF MULTIPLE MELT CONNECTIVITY FROM MANTLE CONDUCTIVITY
Stephen K. Park (IGPP, University of California, Riverside, USA)
Mihai N. Ducea (Dept. Geological Sciences, University of Arizona, Tucson, Arizona, USA)
Models of the effective conductivity of mixed phases (e.g., melt + solid) based on laboratory
measurements are often used to relate electrical conductivity inferred from magnetotelluric studies to
the physical and chemical state of the upper mantle. Rarely do these models permit the inclusion of
multiple phases, however. Hashin-Shtrikman (1962) bounds are commonly used to provide upper
and lower bounds for mixed phases. While rarely used for more than two phases, these bounds do
permit the inclusion of multiple phases. Unfortunately, these bounds do not adequately represent the
connectivity that might exist between multiple conducting phases in the mantle. A new model is
presented here which allows for different connectivities between three phases; this model can be
generalized to many phases, however. We use this model to show that the predicted connectivity
between two conducting phases (sulfide melt and basaltic melt) beneath the Sierra Nevada matches
that observed in laboratory experiments. In particular, sulfide melts form isolated blebs on the
surface of the solid phase and the basalt melt is the interconnected phase. This result suggests that
electrical conductivity may be used to infer in situ melt properties in the mantle.
Hashin, Z., and S. Shtrikman, A variational approach to the theory of effective magnetic
permeability of multiphase materials, J. Appl. Phys., 33, 3125-3131, 1962.
EM1-7
ELECTRICAL CONDUCTIVITY OF WATER-BEARING WADSLEYITE:
IMPLICATIONS FOR THE WATER CONTENT OF THE TRANSITION ZONE
Brent T. Poe (Bayerisches Geoinstitut, University of Bayreuth; now at: Istituto
Nazionale di Geofisica e Vulcanologia,Via di Vigna Murata 605, 00143 Roma)
Claudia Romano (Dipartimento delle Scienze Geologiche, University degli Studi di
Roma Tre, Rome, Italy)
James A. Tyburczy (Department of Geological Sciences, Arizona State University,
Tempe, AZ 85287-1404 jim.tyburczy@asu.edu)
Recent laboratory work has demonstrated that the electrical conductivity profile of the
mantle can be reasonably approximated from the measured conductivities of the major
mineral constituents such as olivine, wadsleyite, ringwoodite and silicate perovskite.
Where the approximation shows lesser agreement with recent geophysical electrical
conductivity models, particularly at transition zone depths, consideration of the effects of
minor constituents that might affect the electrical properties of minerals is required. For
the transition zone, only the conductivities of nominally anhydrous minerals have
previously been measured, yet both wadsleyite and ringwoodite, which compose a
substantial proportion of the transition zone, can incorporate up to several weight percent
water into their structures. Knowing the effect of dissolved water on the electrical
conductivities of these minerals may thus provide some constraints on the amount of
water in the transition zone. We have obtained preliminary results for the electrical
conductivity of hydrous wadsleyite at P,T conditions of the Earth's upper mantle. The
sample material was synthesized at 12 GPa and 1100 C in a welded Pt capsule using a
multianvil apparatus. The recovered material was cut into 0.5 mm thick disks for in-situ
measurement by complex electrical impedance spectroscopy also at 12 GPa in a
multianvil apparatus. Pre-and post-run analyses by secondary ion mass spectrometry
(SIMS) indicate that the samples lost some water during the high P,T conductivity runs.
Post run analyses were 0.40 wt % and 0.50 wt % H2O, respectively for two conductivity
runs performed. We find that the water-bearing wadsleyite is approximately 1/2 to 1
orders of magnitude more conductive than the nominally dry wadsleyite (Xu et al., 2000)
over the temperature range 600 to 1050°C and that conductivity increases with water
content. The activation energies are all similar with values of about 1.0 eV.
Furthermore, analysis of nominally dry wadsleyite prepared in the same manner as that of
the previous workers yields a water content of 0.16 wt % H2O, suggesting that the
previous 'nominally dry' results contained a small amount of water. Based on these
laboratory results, Earth conductivity profiles in the transition zone (420-670 km depth)
can be matched by those of dry wadsleyite, that is, no water is required to exist in the
region.
EM1-8
LABORATORY MEASUREMENTS OF THE ELECTRICAL PROPERTIES OF
GEOTHERMAL RESERVOIR ROCKS FOR RESERVOIR ANALYSIS, FLUID
TRACKING, AND FRACTURE DETECTION
Jeff Roberts (Geophysics and Global Security, Lawrence Livermore National Laboratory,
Livermore, CA, 94551; roberts17@llnl.gov)
Abelardo L. Ramirez, Russ Detwiler, Steve Carlson, Bill Ralph and Brian Bonner
(Geophysics and Global Security, Lawrence Livermore National Laboratory,
Livermore, CA, 94551)
EM geophysical methods are often used in the geothermal industry for exploration,
reservoir evaluation, to track changing reservoir conditions and to optimize fluid
reinjection strategies. The effective use of such geophysical tools requires an
understanding of the physical properties that affect electrical conductivity. Laboratory
measurements of the electrical resistivity of intact and fractured representative
geothermal reservoir rocks were performed to investigate the resistivity contrast caused
by active boiling and to infer saturation and fracture location in a field test. Laboratory
measurements were performed to simulate field test conditions with confining pressures
up to 100 bars and temperatures to 145°C. These measurements are a first step toward
making the search for fractures using electrical methods quantitative. Intact samples
showed a gradual resistivity increase when pore pressure was decreased below the phaseboundary pressure of free water while fractured samples show a larger more abrupt
resistivity increase at the onset of boiling. The change in resistivity for intact samples at
the onset of boiling is controlled by the pore size distribution. The ratio of boiling to
liquid-saturated sample resistivity ranged between 2 and 5. Analysis of a field test
provided the opportunity to evaluate fracture detection using electrical methods at a
larger scale. Interpretation of electrical resistance tomography (ERT) images using
resistivity contrasts determined by laboratory experiments indicates that actively boiling
fractures can be identified. Rapid changes in resistivity as determined by ERT indicated
active fractures and changing fluid pathways. Borehole extensometers showed
simultaneous fracture movement in the area of most rapidly changing electrical
resistivity, corroborating the interpretation.
This work was performed under the auspices of the U.S. Department of Energy by
Lawrence Livermore National Laboratory under contract No. W-7405-Eng 48.
EM1-9
MANTLE CONDUCTIVITY MEASURED IN THE LABORATORY:
A HISTORY
T. J. Shankland (Geophysics Group, Los Alamos National Laboratory,
Los Alamos NM 87545; shanklan@lanl.gov)
Since the electrical conductivity profiles of Lahiri and Price (1939)
indoor geophysicists have attempted to answer the question, what does it
all mean? How do we interpret such profiles of Earth's mantle? More
than half a century has gone into trying to get it right.
As we now know, the basic materials are olivines, pyroxenes,
spinels, garnets, and their high-pressure, high-temperature polymorphs.
However, beginning in the late 1940s, researchers, in the best
geophysical tradition, plunged in by simply measuring conductivities in
ultramafic rocks. As inconsistencies appeared over the next couple of
decades, it was necessary to define minerals in terms of condensed
matter physics—an approach made necessary by the need to extrapolate
to extremes of mantle conditions not available in the laboratory. By
these standards mantle minerals are insulators, and for insulators
electrical transport properties are difficult to measure reliably.
Achieving chemical buffering (principally of oxygen fugacity by
Duba and colleagues) in the early 1970s had two big effects: (1) it threw
into doubt most of the previous quarter-century of work, and (2) it
introduced nearly unprecedented reproducibility. Improved laboratory
measurements permitted the role of iron in charge transfer to be defined
and interpreted in terms of oxygen-sensitive defect populations. For
mantle olivine (~10% fayalite content) there was actually general
agreement among several groups for measurements at mantle
temperatures. [In both field and laboratory conductivity measurements
half an order of magnitude appears to be the level at which
disagreements become academic.]
Attention to chemical buffering has led to other advances. For
instance, measurements of mineral conductivity in multi-anvil devices
and diamond anvil cells have become possible at mantle pressures
and/or temperatures. The role of crystallographic phase transitions was
elucidated. An unusual and surprising agreement with geophysical
observations has been achieved. In another case, "water" in its various
chemical species appears to enhance conductivity, at least in the
uppermost mantle. Elemental carbon could also play a role in this
region. In summary, mistakes as well as successes in this field illustrate
how geophysical science moves toward improved understanding of the
Earth.