Professional Documents
Culture Documents
.. (8-25)
Frequently, an empirical lineas relation between thermal conductivity and density d is used ( for
example cermak, 1976, Cermak and Rybach, 1982)
.. (8-26)
Where a
2,
b
2
are empirically.
As a example of this applicarion figure 8.10. show the results obtained by poulsen et al (1981) for
compact silurian limestones.
Figure 8.19 thermal conductivity vs, density for lemestones (O-dry,- water saturated data after poulson
et al (1981); aquations for the regression curves see text.
An analysis of the data following equation (8-26) result in the relationship for dry rock;
..(8-27)
And for water saturated rock;
.(8-28)
Where the density d is in kgm
-3
and the thermal conductivity in W m
-1
K
-1
Table 8.10 contain some parameters for the relation (8-26) acquired from experimental investigations of
different authors. Kappelmeyer and Haenel (1974) referred to the fact of a high scattering of such
relations because the individual samples differ in its quartz content. Cermak (1967)notes the tendency
of growing valur a
2
with increasing geological age of the sediment.
Table 8.10 empirical coefficients of the equation 8-26 tor sedimentary rocks using data from the
following references (C67-Cermak (1967), C82 Cermak and Rybach (1982) ,H76-Hurtig and Schlosser
(1976), P81- Poulsen (1981).
The equation 8-25 and 8-26 are equivalent; by using the density-porosity relation (equation 3-3) we get
( )
..8-29)
It follows the conversion
)..8-30
Based on a geometric mean model (see section 8.4.2.2) balling et al. (1981) derived by least squares fits
relationships between thermal conductivity and of the from (see equatuon 8-46) for conecoic and
mesozoid sediments (Denmark)
(8-31)
Table 8.11 shows mean values for the parameters a and b
Table 8.11 empirically determined parameters a and b (in W/mK) for equation 8-31 references; B balling
et al., 1981;L-Lovell, 1985 (see also Griffith et al, 1992).
The parameter b is controlled by the matrix properties (see section 8.4.2.2). therefore, it is reatively high
for sandstone and especially for quartz sandstone. An increasing clay content (e.g in case of shaly sands)
results in a decreasing thermal conductivity. A study of this problem- with theoretical consideration-wac
published by Brigaud et al (1989).
Figure 8.9 and 8.10 show the influesce of the thermal propertiesof pore filling material on the thermal
conductivity of porous rock; it is even more distinct for unconsolidated rocks than for consolidated (see
figure 8.9b). somerton (1958) has published an empirical relationship for the ratio of the thermal
conductivity of dry (
)
(
) (
8-32
Where is the porosity and c an empisical exponent in the range of 0.91.3.
The inceasing influence of pore material cinductivity with frowing porosity may also be illustrated by
figure 8.11 using data from kunii and smith (1960; see Roy , 1981) and woodside and messmer (1961),
the ratio of rock conductivity for pores filled with water and air
)(8-33)
Where C
w
is the waer content of the rock in volume fractions. This relation sontains the combined effet
of shanging porosity and changing degree of water saturation (C
w
=S
w
). dachnov and Djakonov (1952)
Djakonov and Jakpvlev (1969) have found a comparable relation with an exponent of 0.8 C
w
.
8.3.3.4 unconsolidated sediments
Some properties and dependencies are discussed in section 8.3.3.3, the following supplements this
discussion.
Marine unconsoidated sediments are a particular type of sediments, which are characterised by high
porosity and very loose bonding between the individual grains. Experimental results have been
publisged by somerton and boozer (1960), Ratcliffe (1960), Hutt and Berg (1968), v. Herzen and Maxwell
(1969), and in the cited data collections.
Figure 8.12 demonstrates the correlation between thermal conductivity and water content (weigh %)
plotted after data from Ratcliffe (1960), and the derived correlation between thermal conductivity and
porosity.
Bullard and Day (1961) suggested for seafloor sediments the empisical relationship between sediment
thermal conductivity and porosity
( ) ( )
..(8-34)
Another special type of unconsolidated rocks are soils. Their thermal conductivity is essentially
influenced by the soil moisture (figure 8.13) and also show a non-linear correlation. This strong
correlation was doscussed by Nibterkorn in 1962
Schuch (1982) may be cited for the general behaviour of soils:The thermal conductivity of dry soils is
small (0.20.8 W/mK), reaches a maximum by 20 to 30 wt% of water content 2 or 3 W/mK), decreases
for higher contents of water, e.g, wet bog, and draws near the value of thermal conductivity of water
(0.6 W/mK). This decrease is originated by an increasing porosity in that range and a connected
decrease of the heat transfer by the skeleton of the solid parts of the soil.
8.3.3.5 dependence of thermal conductivity on pressure and depth
Increasing pressere yields an increasing thermal conductivity in sedimentary rocks (figure 8.14) due
primarily to
Improved heat transport at grain-grain contacts and also at cracks microcracks or other defects
within the compact solid matrix.
Decrease of porosity.
Therefore, pressure conditioned variations of the thermal conductivity are more evident in compressible
rock ( unconsolidated sediments, consolidated sediments woth high porosity) rather than in rocks with
zero or small compressibility (dense carbonates, anhydrite). The direct dependence of thermal
conductivity varations on the deformation behaviour explains the phenomena of nonlinearity and partial
irreversibility (hysteresis) of the thermal conductivity vs. pressure curve.
A log-log presentation of thermal conductivity versus pressure results in a linear correction as a first
approximation. This corresponds to a power law of the form
.(8-35)
Where in this particular case the exponent m is between 0.02 and 0.04.
The depth dependence of thermal conductivity is the combined effect of the pressure and temperature
dependence, of the vertica pressere and temperature gradient and the effects og diagenesis,
compaction and cementation.
Figure 8.15 shows a result for shale samples. The laboratory measurement error is given by the authors
to be less than 5%. The data scatter confirms that the rock structure and the individual composition of
fhe sample (e.g clay and quartz content) is also influential. The general trand was approximated by
different linear and non-linear regressions; where the best approximated by particular case results in
the form.
................................(8-36)
With correlation coefficient of r = 0.82; the depth z is in m.
Figure 8.15 thermal conductivity vs. depth for shales (anjinour depression
The complex effects of depth and the different feolofgical processes results in the tendency of
increasing conductivity with the geological age for comparable rock types. Table 8.11 reflects this for
sedimentary rocks from Germany.
Table 8.11 mean values and standard deviation of thermal conductivity , specific heat c, and density d
of sedimentary rocks from Germany; after Kappelmeyer and Haenel (1974).
Figure 8.16 shows results of sample measurements of unconsolidated sediments trom a borehole in
Zurich Lake (Switzerland). And conductivity increases in the topmost 30 m from about 0.9 to about 1.6
W/mK. From 30 to 137 m the conductivity shows s smaller increase up to about 2.4 W/mK (Finkh, 1983).
The general trend of this experimental data is a nonlinear increase of conductivity with depth.
Figure 8.16 thermal conductivity versus depth determened on cores of unconsolidated sediments from a
borehole in Lake Zurich, Switzerland (after Finckh, 1983).
8.3.3.6 anisotropy of thermal conductivity (sedimentary rocks)
The anosotropy a thermal conductivity correlates with the structural-textural propertios of sedimentary
rocks. Carbonates and sandstone usually show only a small anisotropy (mostly in the range up to 1.3).
higher anisotropy in carbonater (up to about 1.3.1.4)can be originated by a distinct bedding or crack
orientation. Typical anisotropic rocks are shales (up to 2.5) and claystones (up to 4)