close

Вход

Забыли?

вход по аккаунту

?

THE MECHANICS OF HYDROTHERMAL SYSTEMS. Bruce Hobbs

код для вставкиСкачать
THE FUTURE OF
EXPLORATION.
Bruce Hobbs, Alison Ord, John Walshe,
Hans Muhlhaus, Yanhua Zhang, Chongbin
Zhao and Reem Freij Ayoub.
CSIRO Exploration and Mining, Perth, Australia.
CHAPMAN CONFERENCE,
AUGUST 23, 2001.
Structure of this Presentation.
(a) The Problem – We need a new paradigm for
exploration.
(b) A process oriented classification of
hydrothermal mineralising systems.
Isothermal fluid/rock reactions.
Gradient reactions.
Discontinuity reactions.
(c) Processes driving fluid flow.
(d) The modelling software environment.
(e) Some topics for the future.
Self-organisation
The role of mantle dynamics.
(a) The Problem – We
Need a New Paradigm.
The Exploration Industry Has Performed
Poorly
Number of Large Gold + Base Metal Discoveries :
No of Discoveries per year Western World
20
Based on discoveries with
in-situ value >US$1B
The Discovery Rate has remained
flat over the last 30 years
15
10
5
0
1950
1960
From WMC
1970
1980
Page
1990
2000
Sources : WMC Global Deposit Data Base May 01
Metals Economic Group 20004
Number of Large Gold + Base Metal Discoveries :
Western World
Exploration Expenditure (excl Mine
Site)
June 2001 US$B
$4
No of Discoveries per year
20
15
…. but expenditures have
approximately tripled over the same
period
$3
10
$2
5
$1
0
1950
$0
1960
1980
1970
Note : For years prior to 1990 assumes Minesite exploration
makes up 20% of total expenditures
Page
1990
2000
Sources : WMC Global Deposit Data Base May 01
Metals Economic Group 20005
…. Resulting in a Tripling in the Cost per
Discovery
Average Cost per Gold + Base Metal Discovery :
NPV of Voisey’s Bay
Cost per Large Discovery : June 2001 US$m
= US$617m
Western World
$1000
BNP Paribas 2001
Average
in 1990s
US$300m
Average Discovery Costs
have tripled in the last 30 years
$800
$600
$400
Average
in 1950s & 60s
US$100m
$200
$0
1950
1960
Based on discoveries with
in-situ value >US$1B
Average
in 1970s
US$70m
1970
1980
1990
2000
Sources : WMC Global Deposit Data Base May 01
Metals Economic Group 2000
Page
6
The Oil Industry Shows a Similar Trend
Number of Giant Oil Fields : by Discovery Year
1931-94
Number of Discoveries per Year
117
80
Giant defined as
>500 million
barrels
60
40
20
0
1930
1940
1950
1960
1970
The number of giant oil discoveries has dropped
significantly in the last 20 years
1980
1990
2000
Source : Petroconsultants
1998
(b)Modelling of Processes.
as a tool to aid thinking and
explore a range of “what-if”
questions before and during an
expensive exploration program.
This paper is concerned with
the processes that operate in
hydrothermal mineralising
systems.
We are concerned primarily with the
mechanics of these processes.
The term mechanics is used to mean the science
involved in understanding the behaviour of a
fluid saturated porous solid subjected to:
• a general stress state,
• gradients in pore fluid pressure, hydraulic head
and temperature,
and within which chemical dissolution, transport
and reactions may occur.
THE FULLY COUPLED FOUR-FOLD
GENERAL
PROBLEM.
FLOWTHROUGH
OF
CHEMICALLY
REACTIVE
SPECIES
DEFORMATION
LINMKED TO
CHANGES IN
POROSITY AND
PERMEABILITY
STRESS
STATE, sij;
EFFECTIVE
STRESS
INFLUENCED
BY CHANGES
IN PORE
PRESSURE
POROSITY AND
PERMEABILITY
LINKED TO
CHEMICAL
REACTIONS
GRADIENTS IN TEMPERATURE, HYDRAULIC
HEAD AND CONCENTRATIONS OF CHEMICAL
SPECIES
In general there are strong feedback
mechanisms associated with these
processes, so that each process has
an influence upon the others, and
part of our goal is to take these
feedback processes into account in
a quantitative manner.
The four-fold, fully coupled system involved in
mineralising systems.
(b) A Process Oriented
Classification of
Hydrothermal Mineralising
Systems.
Mineralisation in hydrothermal
systems arises from one or a
combination of the following three
fundamental processes:
(i) Isothermal fluids/rock reactions,
(ii) Gradient reactions,
(iii) Discontinuity reactions.
The three end-member types of
hydrothermal ore bodies.
GRADIENT
DISCONTINUITY
FLUID-ROCK REACTION
(i) Isothermal fluid-rock
reactions.
G
D
F/R
Isothermal Fluid Rock Reactions.
Rock type A
Darcy
velocity uA
Equilibrium
concentration cA
Reaction
Front
Rock type B
Darcy
velocity uB
Equilibrium
concentration cB
Mineralisation rate proportional to (cA – cB).
Reaction front velocity proportional to uB.
Thus, for isothermal fluid-rock
reactions,
• The grade is proportional to (cA-cB),
• The tonnage is proportional to the
Darcy fluid velocity, uB.
Characteristics of Ore Bodies
with Dominant Origins through
Fluid/Rock Reactions.
• Isothermal fluid/rock reactions are essentially
replacement processes where the extent of the ore
body is controlled by the magnitude of the Darcy
fluid velocity.
• Such ore bodies can be of exceptionally high grade
but are patchy in their development since very small
changes in permeability can lead to fluid focussing
leaving low permeability rocks barren.
• Examples are the Irish Pb/Zn and Hamersley Fe
deposits.
(ii) Gradient Reactions.
G
D
F/R
Mineralisation due to fluid flow down a
gradient in equilibrium concentration.
Rate of mineralisation = -fu.grad ce
Darcy Fluid
Velocity, u
Porosity,
Porosity,
f. Porosity,
f.
f.
Gradient of equilibrium concentration, grad ce.
Mineralisation due to fluid flow obliquely across a
gradient in equilibrium concentration.
100 ppb
10 ppb
Darcy
Fluid
Velocity, u
Darcy Fluid
Velocity, u
Porosity, f.
1 ppb
Rate of mineralisation = -fu.grad ce
Mineralisation Rate with no Local
Chemical Reaction.
• Mineralisation rate = -fu.grad ce
• Since ce = f ( T, p, cr ),
• The Chain Rule of differentiation gives
us:
ce
 ce
Q m пЂЅ пЂ­f u п‚· пѓ­
пѓ‘T пЂ«
пѓ‘p пЂ«
p
 T

r
пѓј
пѓ‘ cr пѓЅ
cr
пѓѕ
ce
Mineralisation Rate with Local
Chemical Reaction.
• Mineralisation rate = -fu.grad ce + Ri
Or:
ce
 ce
Q m пЂЅ пЂ­f u п‚· пѓ­
пѓ‘T пЂ«
пѓ‘p пЂ«
p
 T

r
пѓј
пѓ‘ c r пѓЅ + Ri
cr
пѓѕ
ce
PROGRESS OF REACTION
FRONT IN FLUID ROCK
REACTION.
H2O+CO2
K-SPAR+QUARTZ
100 YEARS
5000 YEARS
K-SPAR
MUSCOVITE
PYROPHYLLITE
QUARTZ
Many gradients in equilibrium
concentrations of metals arise
from fluid mixing.
An important mechanism that assists fluid
mixing is the focussing of fluid flow into
regions which have high permeability
relative to their surroundings.
Fluid focussing into high permeability
lenses.
For long, thin lenses with high permeability
contrasts the focussing effect is dramatic.
Concentration gradients are also
dramatically changed by focussing.
FLUID
FOCUSSING
IN A NARROW
LENS.
Vertical Darcy
velocity 8*10-8ms-1.
STREAMLINES IN A NARROW LENS.
GOLD GRADES IN A NARROW LENS.
S tilfo ntein
facies
LT S zon e
NW
No C0
“B uffer” fluid
faults
Zaaiplaats facies
Lo w Tensile
S tren gth zone
2
C H 4 - C 0 2 fluid
decom p ressio n
m ixing zon e
Zuiping
C 0 2 fluid
4
4
Je
rs
ey
Reef X Section Geology
Permeability Structure
CO2
CH4
CH4 Concentration
CO2 Concentration
Au Precipitation
MAXIMUM RATE OF MINERALISATION:
20 g/tonne/million years
U Precipitation
MAXIMUM RATE OF MINERALISATION:
100 g/tonne/million years
Aqueous species at 0.025 million years
Gold
Chlorite
Pyrrhotite
Muscovite
Pyrophyllite
Graphite
K-FELDSPAR
Redox
vs
pH
Diagram
Witwatersrand - Basin Scale Alteration
K-FELDSPAR
Redox
vs
pH
Diagram
Characteristics of Ore Bodies
with Dominant Origins through
Gradient Reactions.
• Deposits are characteristically of high
tonnage but rarely are of bonanza grades.
• Since grade is proportional to Darcy fluid
velocity, mineralisation can be quite
extensive.
• Examples are Witwatersrand gold and
much of the gold mineralisation in the
Yilgarn of WA.
(iii) Discontinuity Reactions.
G
D
F/R
Discontinuity Reactions.
• Here a change in pressure and/or
temperature lead to boiling or to other
forms of phase separation; H2S may be
bled from the system or salinity may be
reduced.
• This leads to changes in salinity, pH
and/or Eh and hence changes in the
equilibrium solubility of metals.
Discontinuity Reactions.
Equilibrium
concentration, c
Equilibrium
concentration, ce
Mineralisation rate, and hence the grade,
proportional to (c-ce).
Tonnage governed by size of dilating region.
If there is a pore pressure drop in this system
there is an extra component to the
mineralisation rate:
{
Mineralisation Rate = - u.
G
ce
p
}
пѓ‘p
Many mineralised vein systems
are probably hybrids like A.
Notice, no need for seismicity.
D
F/R
A model with a
single fault
4 km
Higher shear stress
lower shear stress
Dilation zone arrays are
developed at fault tips
Red to Green – volume increase
Adding fluid
sources (flux) at the
top and bottom of
the model
A major dilation
zone developed
between the two
faults
Plot of flow
stream lines,
showing the
mixing of two
fluids, from
different
sources
A model with
multiple
faults/fractures
and fluid sources
Flow stream
lines, showing
fluid flow and
mixing patterns
Permeability
structures
reflect damage
zones (tensile
failure) related
to individual
faults
1. Dilatant areas – structurally
favorable locations
2. Fluid focusing/mixing – high
flow rates
3. Areas with greater concentration gradient of important
chemical species (e.g. CH4) and areas of chemical reaction front
Characteristics of Ore Bodies
with Dominant Origins through
Discontinuity Reactions.
• Deposits can be of exceptionally high grade but
are of limited spatial extent since the geometry
of the system which produces the discontinuity
in metal equilibrium solubility, such as a
pressure drop, is commonly quite localised
spatially.
• Examples are many vein hosted gold deposits
such as Bendigo-Ballarat.
(c) Processes Driving Fluid
Flow.
TOPOGRAPHICALLY
DRIVEN FLOW.
LOCAL
DEVOLATILISATION.
EXTERNALLY DERIVED
FLUIDSDEVOLATILISATION OR
IGNEOUS.
HYBRID.
(i) Topographically driven
fluid flow.
• Widely used as important fluid driving
mechanism for MVT deposits; eg Grant
Garven and co-workers.
• Undoubtedly an important mechanism
but there are aspects we still do not
understand.
• In particular, there seems to be a mass
balance and timing problem.
Century fluid system supplied by “squeegying” of deep basin
aquifers Hydraulic head promoted by meteoric water input
(height of orogen approx 2-3km above Century position,
giving effective hydraulic head of ~500m-1km?)
Meteoric water drive interrupted by structuring, Century
fluid cell declines
Inspired by Oliver (1986) and Garven & Freeze (1984)
Fluid flow vectors associated with critical faults and
stratigraphic units in the Century system
The Problem with
Topographically Driven Flow.
Consider the following situation:
B
A
100km
A Darcy flow rate of 1 m y-1 is equivalent to 1 m3 m-2 y-1.
Thus a cubic meter of fluid at A would be transported to B in
105 years.
(ii) Dilatancy driven fluid
flow.
Often thought of in terms of “fault valve
pumping” (eg., Sibson). However, seismic
behaviour is not essential here and any
dilatancy associated with deformation,
distributed or localised, will lead to a
gradient in hydraulic head to drive fluid
flow.
(iii) Fluid flow driven by
devolatilisation.
Typical metamorphic fluxes as recorded by
metamorphic petrologists are in the order of 10-9
kg m-2 s-1 ( eg., Connolly).
At porosities and permeabilities to be expected in
metamorphic rocks, this corresponds to Darcy
velocities of say 10-7 m s-1.
Although such fluids are undoubtedly important in
mixing to change pH and Eh, their volumes do
not seem large enough to transport significant
masses of metal.
TWO BASIC MECHANISMS.
• HYDROFRACTURE: if the rate of fluid
pressure generation rate exceeds the
dilatancy rate.
• POROSITY WAVES: if the dilatancy
rate exceeds the rate of fluid pressure
generation.
THE ORIGIN OF POROSITY WAVES.
LITHOSTATIC
GRADIENT
HYDROSTATIC
GRADIENT
FLUID PRESSURE
DEPTH
DEPTH
TENSILE FAILURE
PORE COLLAPSE
FLUID PRESSURE
Thus if we look at the various
mechanisms for transporting significant
masses of metals:
Topographic: A problem.
Devolatilisation: A problem.
Igneous sources: OK.
Convection: Let us now look.
(iv) Fluid flow driven by
thermal convection.
FLUID FLOW VECTORS
STREAMLINES
TEMPERATURE
VELOCITY
H2S
Galena
STREAMLINES
SO42-
Sphalerite
TEMPERATURE
H+
Pyrite
CHARACTERISTICS
OF THERMAL
CONVECTION
SYSTEMS WITH
FIXED TOP AND
BOTTOM
TEMPERATURES.
It has long been claimed (eg.,
Wood) that convection is impossible
in a system with a lithospheric fluid
pressure gradient.
• This is true for a system with impermeable
upper and lower boundaries and fixed
temperatures at these boundaries.
• However, for other boundary conditions, this is
not true.
• For example, with fixed pressure and thermal
flux boundaries, convection is possible with
vertical fluid fluxes.
VARIATION OF CRITICAL RAYLEIGH NUMBER
WITH MAGNITUDE OF UPWARD FLOW.
20
Ra
critical
0
0
Pe
6
Thermal convection in a system with
lithospheric pressure gradient.
Flux due to
lithospheric pore
pressure gradient.
Upper
Boundary:
constant
pressure and
constant
thermal flux
Lower
Boundary:
constant mass
and thermal
fluxes
Thus, systems such as shown below, deep or shallow in the
crust, seem to be one of the few ways of transporting
significant masses of metal over extended periods of time.
Impermeable Base & Top
Convective temperature
CO2
Permeable Fault,impermeable boundaries
L
ko
10 ko
ko
L
3L
Flow focussed
in the fault
Convective T &
Vg
Zero convective pressure at
boundaries
Convective Temperature
Geological
Model
In
GOCAD;
Imported
from
standard
CAD
package.
Convective Temperature
CO2
(d) The Modelling Software
Environment.
Model in FLAC3D.
Coupled FLAC-FASTFLO
Coupled FLAC-FASTFLO
Coupled FLAC-FASTFLO
Gold bisulphide
Graphite
Sketch of Program Interfaces
Geological model
(e.g. in Gocad)
Return to Gocad
for
3D visualisation
Control of
simulation via
browser file (XML)
Process 1
(e.g. Flac3D)
Process 2
(e.g. FastFlo)
(e) Some Topics for the
Future.
Mantle Dynamics.
DISTRIBUTION OF METALS
THROUGH TIME.
• Most gold on Earth was emplaced between 2.8
and 2.6 billion years ago.
• Most lead-zinc on Earth emplaced 1.7 billion
years ago.
WHAT GLOBAL SCALE PROCESSES WERE
RESPONSIBLE FOR THESE
MINERALISING EVENTS ?
Outcomes from modelling mantle
behaviour have been very
significant in the past decade.
These outcomes include:
• Convection is dominated by hot upwelling
plumes and cold down going sheets.
• Temperature increases due to viscous
dissipation are important on the edges of slabs
and plumes.
Further Outcomes.
• Stratified mantle convection exists at high
Rayleigh Numbers (Ra) and whole mantle
convection at low Ra.
• Progressive cooling of the Earth’s core,
together with nonlinear effects associated with
radioactive internal heating leads to massive
flow instabilities and large increases in thermal
flux through the lithosphere.
The Future.
• Modern seismic tomography of the Earth is
defining the details of modern thermal
convection patterns.
• Modelling will enable us to describe past
convection systems and define periods of
convective instability or periods of gross change
in behaviour.
• Signatures of these changes in the geological
record will enable us to identify these effects at
the surface and understand why some periods
of time and some regions are more endowed
than others.
The Beginning.
Thank you.
Документ
Категория
Презентации
Просмотров
6
Размер файла
13 502 Кб
Теги
1/--страниц
Пожаловаться на содержимое документа