SPECIAL ISSUE
Petrographical and petrophysical properties of sandstones:
statistical analysis as an approach to predict material
behaviour and construction suitability
H. Stu¨ck • R. Koch • S. Siegesmund
Received: 10 June 2012 / Accepted: 12 September 2012 / Published online: 25 October 2012
 The Author(s) 2012. This article is published with open access at Springerlink.com
Abstract Most studies dealing with material properties of
sandstones are based on a small data set. The present study
utilizes petrographical and petrophysical data from 22
selected sandstones and *300 sandstones from the litera-
ture to estimate/predict the material and weathering
behaviour of characteristic sandstones. Composition and
fabric properties were determined from detailed thin sec-
tion analyses. Statistical methods applied consist of data
distributions with whisker plots and linear regression with
confidence regions for the petrophysical and weathering
properties. To identify similarities between individual
sandstones and to define groups of specific sandstone types,
principal component and cluster analyses were applied.
The results confirm an interaction between the composi-
tion, depositional environment, stratigraphic association
and diagenesis, which leads to a particular material
behaviour of sandstones. Three different types of pore radii
distributions are observed, whereby each is derived from
different pore space modifications during diagenesis and is
associated with specific sandstone types: (1) bimodal with
a maximum in capillary and micropores, (2) unimodal
unequal with a maximum in smaller capillary pores and (3)
unimodal equable with a maximum in larger capillary
pores. Each distribution shows specific dependencies to
water absorption, salt loading and hygric dilatation. The
strength–porosity relationship shows dependence on the
content of unstable lithic fragments, grain contact and type
of pore radii distribution, cementation and degree of
alteration. Sandstones showing a maximum of capillary
pores and micropores (bimodal) exhibit a distinct hygric
dilatation and low salt resistance. These sandstones are
highly immature sublitharenites–litharenites, characterized
by altered unstable rock fragments, which show pointed-
elongated grain contacts, and some pseudomatrix. Quartz
arenites and sublitharenites–litharenites which are strongly
compacted and cemented, show unimodal unequal pore
radii distributions, low porosity, high strength and a high
salt resistance. The presence of swellable clay minerals in
sublitharenites–litharenites leads to a medium to high
hygric dilatation, whereas quartz arenites show little hygric
dilatation. Sandstones with unimodal equal pore radii dis-
tribution mostly belong to weakly compacted and cemen-
ted mature quartz arenites. These are characterized by high
water absorption and high porosity, low to medium
strength and a low salt resistance. The data compiled in this
study are used to create a sandstone quality catalogue.
Since material properties are dependent on many different
parameters of influence, the transition between different
lithotypes is fluent.
Keywords Sandstones  Pore space  Prediction 
Weathering behavior  Compressive strength
Introduction
Due to their great availability, sandstones are commonly
used as a natural stone for buildings, monuments and
sculptures. The wide variety of different applications
increases the need to understand their characteristic mate-
rial properties. Complex sedimentary processes are
involved in the origin of sandstones. These include the
H. Stu¨ck (&)  S. Siegesmund
Geoscience Centre of the University Go¨ttingen,
Goldschmidtstr.3, 37077 Go¨ttingen, Germany
e-mail: hstueck@gwdg.de
R. Koch
University of Erlangen, GeoCentre Northern Bavaria,
Lo¨wenichstr.28, 91054 Erlangen, Germany
123
Environ Earth Sci (2013) 69:1299–1332
DOI 10.1007/s12665-012-2008-1
source material, transport processes, the depositional
environment and diagenesis. Characterizing and predicting
the material properties of sandstone requires detailed
investigations of sedimentological, petrographical and
petrophysical parameters (Koch and Sobott 2008; Morales
Demarco et al. 2007; Ru¨drich et al. 2010). Thereby, the
characteristics of the pore space such as porosity, permeability
and pore geometry are most important for the weathering
behaviour of porous sandstones (cf. Fitzner 1988; Fitzner and
Snethlage 1982; Putnis et al. 1995; Putnis and Mauthe 2000;
Ru¨drich et al. 2010). In this context, a linear correlation
between porosity and compressive strength and flexural
strength as well as water balance and storage properties has
been reported by Ru¨drich et al. (2010).
However, most of these studies are based on a small data set
of samples and parameters. Therefore, general statements
covering a wide stratigraphic range and different sandstone
types of varying depositional environments and diagenetic
development cannot be made. A statistical analysis of the
parameters such as grain size, compositional and textural
maturity, cementation and fabric, which essentially control
the material properties of sandstones, is lacking.
Therefore, the aim of the present study is to establish
criteria and systems, which aid in predicting the weathering
behaviour of various sandstone types. This is based on a
statistical analysis of a large data set that includes most of
the parameters generally used when analysing the quality
of a sandstone. The study is based on a large amount of
petrographical and petrophysical data collected from 300
sandstones described in the literature. Twenty-two of the
total samples analysed represent important sandstones used
in Germany. These sandstones were investigated in detail
by petrographic analysis, determining the fabric charac-
teristics, measuring the petrophysical properties and com-
bining all these data using statistical analyses.
When using data from the literature, it has to be taken
into account that the data might not be complete in regard
to petrographical analyses and petrophysical measure-
ments. Commonly, just single samples are taken when
analysing a quarry, and thus the samples collected may not
represent the wide variety of facies types present. For
sandstones frequently used, this might be less problematic
because only certain qualities are generally regarded.
Further possible sources of error may be caused by dif-
ferent approaches during sampling, preparation of samples,
different methods and the apparatus used for analysis (Rac
and Chernyshev 1970). Data found on the Internet or from
company brochures need to be used with caution, since the
information may only reflect the optimum quality but not
the natural variations of quality in a quarry. Therefore, the
data utilized in the present study originates from scientific
non-commercial sources, which deal with material testing
and/or research.
Sedimentary processes as a key to understanding
material behaviour of sandstone
The path from unconsolidated sediment to lithified sand-
stone suitable for construction is controlled by complex
sedimentary and diagenetic processes (Tucker 2001). The
mineralogical composition of sandstone (compositional
maturity) and the textural maturity are dependent on sed-
imentological parameters (e.g. transport, depositional
environment), which contribute to the petrophysical prop-
erties, and to a certain degree the weathering behaviour.
Diagenesis begins immediately after/with deposition and
includes processes of compaction, pressure solution,
recrystallization, dissolution, replacement/alteration, clay
mineral and feldspar authigenesis as well as cementation
by varying mineral phases (Burley et al. 1985; Morad et al.
2000; Tucker 2001; Koch and Sobott 2008). Diagenetic
realms include eo-, meso- and telogenetic processes
(Choquette and Pray 1970) acting from shallow to deep
burial conditions depending on the chemistry of pore water
present. The development of pore space geometry through
time together with the alteration of grain contacts are
parameters of special importance. Cements and compaction
lead to the reduction of the primary intergranular pore
space. Eogenetic cements also shelter the rock against later
compaction during deeper burial (Worden and Morad
2000). Some cement is dissolved by processes forming a
late-stage secondary porosity.
Clay material as a primary matrix, pseudomatrix and/or
later infiltration into open pore spaces is a significant
problem of sandstones with regard to their weathering
stability, since they considerably influence porosity and
permeability (Houseknecht and Pittman 1992). Neofor-
mation, transformation and alteration of different clay
minerals under varying conditions (temperature, pressure
and water access) define clay minerals as a special
parameter in sandstones that requires careful analysis when
characterizing the material behaviour of a sandstone.
Methods
To support this investigation, a large data set was taken
from the literature (Fig. 2; Table 1) dealing with the pet-
rographical and petrophysical analysis of sandstones
(Bartelsen 2008; Seidel 2004; Chitsazian 1985; David
2006; Fitzner 1970; Grimm 1990; Holzwarth 1996; Kirsten
2009; Mosch and Siegesmund 2007). Basic parameters
such as porosity, density, water uptake and strength are
included for the sandstones from all over the world, par-
tially originating from the natural stone industry and from
the data collected in detailed studies dealing with German
sandstones.
1300 Environ Earth Sci (2013) 69:1299–1332
123
Petrographical and petrophysical properties of 22
selected sandstones representative for sandstone types from
various geological periods were analysed. Standard
parameters were measured according to the DIN standards
(German Industry Norm) for material testing.
Material parameters measured according to international
standards (e.g. DIN) were used without any further adap-
tation. In contrast, the petrographic data had to be adapted
in some cases for the development of a uniform classifi-
cation. Data quality for grain size and sorting is different
depending upon whether detailed analyses are made or just
estimations are given. Furthermore, not for every rock each
petrophysical property could be found in literature.
Therefore, the number of samples shown in the diagrams
(see Chap. 3.2) can vary.
Petrographical analysis
Thin section analysis was carried out on the 22 represen-
tative samples. Classification was done by point counting
(Van der Plas and Tobi 1965) and using the classification
system of McBride (1963), which allows integrating the
weathering sensibility of sandstones (Hayes 1984). Fu¨cht-
bauer (1988) counts chert and quartzite fragments as a
group of rock fragments, whereas in the classification of
McBride both are interpreted as stable grains and thus
belong to the quartz group. Therefore, in the classification
system of McBride, two weathering-sensitive end members
and a weathering-resistant one are differentiated.
In the classifications of Gilbert (1954) and Pettijohn
(1975), the matrix is generally considered, which is useful
for rocks with a matrix content [15 %. Thus, (grey-)
wackes are seldom used for construction because of their
unsuitable weathering properties and are therefore not
considered in the present study. According to this, data
from the classification of sandstones investigated by David
(2006), Grimm (1990) and Fitzner (1970) were transferred
to McBride’s classification.
Fabric parameters determined by thin section analysis
are grain contacts, sorting (Pettijohn et al. 1987), roundness
as well as textural and compositional maturity. Grain size
analyses determined by thin section do not completely
correspond to grain sizes measured by sieving as discussed
by Mu¨ller (1964).
The visible intergranular porosity and amount of
cements were estimated to acquire information about the
compaction history. The data were plotted in the ‘‘House-
knecht diagram’’ (Houseknecht 1992). Besides the classi-
fication of maturity by McBride (1963), the varying
degrees of alteration, reflecting increasing alteration of
unstable components such as feldspar and lithoclasts, were
subdivided into the classes (1) high, (2) moderate, (3) little
and (4) no degree of alteration.
X-ray diffractometry (XRD) and X-ray fluorescence
were applied to determine the bulk and clay mineral
composition of selected sandstone samples. The cation
exchange capacity for crushed whole rock samples was
analysed using Cu(II)-triethylene tetramine. X-ray dif-
fraction was done for grain size fractions of\2 lm. Grain
size fractions \2 lm were collected from fractions
\63 lm by differential settling in distilled water, i.e. by
using the Atterberg method following Stoke’s law. The
collected sample is prepared by placing them onto ceramic
plates, air dried and coated with ethylene glycol. The
samples were measured at 2H = 2–35 with 0.02 and 3 s
per step.
Petrophysical analyses
Petrophysical parameters including data on the pore space,
strength as well as water transport and storage were
determined according to the DIN standards. Porosity,
matrix density and bulk density were measured in accor-
dance with DIN 52102. The dry mass was determined on
samples after 2 days of preconditioning at 20 C and 15 %
relative humidity. The hydrostatic and the saturated mass
were detected after water saturation under vacuum.
Pore radii distribution was accomplished on cylindrical
samples (10 mm in diameter, ca. 20 mm in length) using
mercury porosimetry (van Brakel et al. 1981; porosimeter
2000 Carlo Erba). The pressures used up to 2 kbar allow
the evaluation of pore radii of about 0.005 lm.
The capillary water absorption was measured according
to DIN 52103 with an under-floor balance, parallel and
perpendicular to the bedding. The measurements were
performed on sample cubes (65 9 65 9 65 mm), which
were dipped to 5 mm in water. The weight increase was
digitally measured every 20 s. The capillary water absorp-
tion can be expressed by the water absorption coefficient
(w value), which represents the absorbed water amount
dependent on area and time (kg/(m2*Hh)). The saturation
degree was calculated from the water uptake under atmo-
spheric and vacuum conditions on the same samples.
Water vapour diffusion was measured with the wet-cup
method (DIN 52615) on samples 40 mm in diameter and
10 mm in length. Samples were exposed at 30 C and a
relative humidity of 50 %. The weight was measured four
times every 24 h. The sorption/desorption was measured in
a climate chamber during a stepwise increase of moisture
between 15 and 95 % relative humidity. A stepwise
decrease was performed until the humidity reached the
initial level of 15 %. Every humidity level was held for
24 h. After equilibration, the mass of the samples was
weighed.
The tensile strength was measured using the Brazil test
according to DIN 22024. The uniaxial compressive
Environ Earth Sci (2013) 69:1299–1332 1301
123
strength was measured on cylindrical samples (50 mm 9
50 mm in size) with coplanar end faces (accuracy of
0.1 %) in the dry and water-saturated state. The load was
applied to the end faces of the specimen with a strain rate
of 1,000 N/s until failure.
Analysis of weathering properties
The hygric dilatation was determined in dependence of the
water saturation. In general, the expansion is determined
with an incremental displacement transducer (resolution
about 5 lm) on samples of 10 cm in length and 15 mm in
diameter. A preconditioning of the specimens was achieved
by drying in a climatic chamber at 20 C and 15 % relative
humidity.
Salt splitting tests were accomplished following DIN
52111 with a 10 % solution of sodium sulphate. The
sample cubes (65 mm edge) were soaked for 6 h in solution
and afterwards dried for 12 h in a drying chamber at 60 C.
After this, they were weighed and photographically docu-
mented. The experiment was ended after reaching 100
cycles or after a loss of 40 % of the original material. The
salt weathering resistance of sandstones analysed as well as
data from David (2006) were grouped as follows:
Group A = no material loss until the 100th cycle
Group B = 50 % material loss until 30–40 cycles
Group C = lose 50 % of their weight after
the 15th cycle.
Statistical methods (univariate, bivariate
and multivariate)
Statistical analyses can be divided into univariate, bivariate
and multivariate methods and each method considers a
different dimensional problem. The statistical treatment of
the data set was performed with the open software ‘‘R
GUI’’.
To analyse the distribution of one rock property (uni-
variate), box-whisker plots were used as shown in Fig. 1a).
The data are illustrated in a ‘‘five-point-summary’’, which
includes the median, two quartiles, extreme values and
outliers. The box represents the area where 50 % of the
data are included and which is limited by the upper (75 %
quartile) and lower median (25 % quartile). The length of
the box expresses the so-called interquartile ranges (IQR)
and is also a degree of scattering. The median divides
the total diagram into two parts, and its position allows
an evaluation concerning the skewness of the data
Fig. 1 a Whisker plot showing
mean, lower and upper extreme
data as well as outliers of
univariate data, b possible
distribution of data inside a
regression analysis with
regression coefficient, and
c example for regression
analyses between water uptake
and porosity with confidence
region of 80 % probability
range, illustrated by the blue
ellipsoid. d Diagram of the
multivariate method showing
the principal component
analyses: rock properties
present the variables and are
illustrated by blue axes. Arrows
pointing in the same direction
are dependent on each other;
arrows pointing in the opposite
direction show negative
correlation, whereas arrows that
perpendicular to each other are
independent of each other. For
further explanation, see text
1302 Environ Earth Sci (2013) 69:1299–1332
123
distribution. According to Tukey (1977), the length of the
whiskers is limited by 1.5 times the interquartile range. All
values outside of the whiskers are characterized as outli-
ers. In the present study, this method was used for all data
of individual properties as well as for recognizing inter-
relationships between the composition and age of the
sandstones.
To investigate the relationship between two properties,
essentially a two-dimensional problem, a linear regres-
sion was performed (bivariate method). The values of two
properties are represented in correlation point diagrams.
The correlation coefficient expresses the relationship of
two features (Bravais–Pearson coefficient). The coeffi-
cient can vary between ?1 and -1; thereby, a positive
value describes a linear relationship with a positive
increase (Fig. 1b i) and a negative value a relationship
with a negative increase (Fig. 1b ii). Correlations with no
linear relationships, e.g. a cloud arrangement of points,
are shown in Fig. 1b iii. Furthermore, within regression
analyses confidence regions with 80 % probability range
are determined under the assumption that basic properties
like porosity, density or strength show normal distribu-
tion. Exemplarily, the relationship between water
absorption and porosity (Fig. 1c) is illustrated, where
80 % probability range is presented by a small, elongated
ellipse.
To combine the relationships of all the measured
parameters, which are important for characterizing the
material behaviour of sandstone, multivariate methods of
principal component and cluster analyses were performed.
Principal component analyses try to reduce a multidimen-
sional problem. The hypothetical variables are constructed,
which minimize the entire sum of squared deviations of a
best-fit straight line. The calculated components try to
describe the wide difference/variance between the proper-
ties and the samples. Since the data of sandstones are based
on different physical units, they differ in dimension and the
influence of properties with high variance can overlay the
variance of other properties, which would lead to an
incorrect calculation of components. Therefore, principal
component analyses were accomplished after previous
standardization. The pre-treatment includes a calculation of
the q–q-norm and a correlation matrix (Dohrmann and
Ku¨hn 2009). Where necessary, data were logarithmized.
Principal component analyses are illustrated with biplots,
which include information on variables (rock properties)
and samples in one diagram (Fig. 1d). The position of
variables to each other is expressed by the loadings, which
are calculated by the variances of the variables. They are
illustrated as arrows, and the position to each other
expresses the particular relationship (Fig. 1d). Arrows
pointing in the same direction have a positive relationship
to each other. In contrast, arrows pointing in opposite
directions exhibit a negative relationship to each other. If
there is no relationship between two properties, arrows
stand perpendicular to each other. Samples are depicted as
points and the coordinates within the diagram in this
multidimensional system are calculated via scores, which
are based on the loadings of the components.
Cluster analyses were done for identifying groups of
similar sandstones, based on all known properties. The
principle of cluster analysis follows two steps: (1) deter-
mination of similarities between the objects and their
properties, and (2) determination of the cluster number.
The similarity is determined via proximity degree and is
calculated with the Euclidian distance matrix. For illus-
tration, a so-called dendrogram was used.
Results
The petrographic analysis includes all the data from this
study as well as from the literature. Both reveal a great
variety of sandstone properties based on data from different
stratigraphic levels.
General remarks on sandstones from different sources
and stratigraphic position
Most of the data originate from Middle European sand-
stones (Fig. 2), which were predominantly deposited in a
cratonic basin. For a discussion of the petrography and
material properties, the stratigraphic position must also be
taken into account as documented by Grimm (1990),
Chitsazian (1985), David (2006), Fitzner (1970), Holz-
warth (1996), Kirsten (2009), Seidel (2004), Bartelsen
(2008) and Koch and Sobott (2008).
The occurrence of different sandstones in the strati-
graphic column is grouped according to their importance
and regional distribution. In Germany, the most important
sandstones occur in the Lower and Upper Triassic (Fig. 2).
Ten sandstone quarries were evaluated and analysed in the
present study. Lower Triassic sandstones are quarried at the
Weser River, in the Hessische Senke, in Spessart, Oden-
wald, in the Pfa¨lzerwald Mountains and in the Black For-
est. Upper Triassic sandstones are predominantly mined in
the Thuringian Basin, at the River Main, and between the
Black Forest and the Swabian Alb. Sandstones, Cretaceous
in age, are also significant (six quarries evaluated), which
are concentrated in the ‘‘Mu¨nsterla¨nder Becken’’, the Hil-
desheim Bay in northeastern Germany and at the Elbe
River in eastern Germany. Carboniferous and Permian
sandstones are rare due to their low occurrence. They are
mined in the north of the Harz Mountains, at the Thurin-
gian–Bavarian border, Hunsru¨ck and in the north of the
Rhenish Slate Mountains.
Environ Earth Sci (2013) 69:1299–1332 1303
123
The sediments of the geologic ages considered are the
result of changing depositional environments through time,
associated with transgressive and regressive marine influ-
ence. The depositional environments of the sandstones
range from pure clastic terrigeneous (fluviatile, aeolian) to
marine facies, which are also reflected in their character-
istic sedimentary structures. Most of the sandstones can be
classified as quartzites, lithic subarkoses, sublitharenites as
well as minor litharenites and subarkoses. The distribution
of sandstone types is characteristic for epicontinental
basins such as the German Basin. Therefore, arkoses are
represented only in very minor amounts. Short descriptions
of different depositional facies are given, which change in
each sedimentary basin through time.
Carboniferous are the oldest sandstones analysed in the
present study (e.g. the Ibbenbu¨ren sandstone). The sedi-
ments were deposited in flood plains and alluvial fans to
the north of the Variscan orogeny (Katzung 1975; Plein
1993). They consist of feldspathic litharenites and sublit-
harenites and partially contain conglomeratic layers and
coal particles. The Ibbenbu¨ren sandstone is characterized
by strong compaction and cementation and also a clayey
matrix.
During the Permian, the German Basin was filled with
shallow marine to terrestrial mixed carbonate/siliciclastic
sediments (Aigner and Bachmann 1992). The Rotliegend
sandstones crop out in southern Germany, Thuringia and at
the Flechtinger mountain range, where terrestrial deposits
Fig. 2 Map of quarries and
sandstones investigated.
Outcrops of deposits from
Carboniferous, Permian, Lower
and Upper Triassic as well as
Cretaceous are illustrated.
Sample locations and literature
sources for different sandstones
are indicated. Numbers
represent the sandstones
investigated in this study
(Table 1)
1304 Environ Earth Sci (2013) 69:1299–1332
123
T
a
b
le
1
P
et
ro
g
ra
p
h
y
an
d
fa
b
ri
c
o
f
al
l
2
2
sa
n
d
st
o
n
es
an
al
y
se
d
S
am
p
le
C
la
ss
ifi
ca
ti
o
n
M
cB
ri
d
e
(1
9
6
3
)
Q
tz
/F
sp
/
L
it
h
o
C
em
en
t/
m
at
ri
x
R
o
u
n
d
n
es
s/
so
rt
in
g
G
ra
in
co
n
ta
ct
M
at
u
ri
ty
&
d
eg
re
e
o
f
al
te
ra
ti
o
n
,
cl
ay
m
in
er
al
s,
ro
ck
fr
ag
m
en
ts
C
re
ta
ce
o
u
s
U
p
p
er
C
o
tt
a
(C
O
)
Q
u
ar
tz
ar
en
it
e
9
5
/5
/0
C
la
y
ey
m
at
ri
x
S
u
b
an
g
u
la
r/
w
el
l
E
lo
n
g
at
ed
L
o
w
er
m
at
.
th
ro
u
g
h
cl
ay
ey
m
at
ri
x
&
al
te
re
d
fs
p
,
g
la
u
k
o
n
it
e
R
u¨
th
en
(R
U¨
)
Q
u
ar
tz
ar
en
it
e
9
0
/1
0
/0
F
ew
sy
n
t.
Q
tz
/
k
ao
li
n
it
e,
il
li
te
S
u
b
ro
u
n
d
ed
/
w
el
l
E
lo
n
g
.c
o
n
ca
v
e-
co
n
.
H
ig
h
,
m
at
u
ri
ty
,
g
la
u
co
n
it
e
L
o
w
er
O
b
er
n
k
ir
ch
en
(O
B
)
Q
u
ar
tz
ar
en
it
e
9
5
/3
/2
H
ig
h
am
o
u
n
t
sy
n
t.
q
tz
/
cl
ay
S
u
b
an
g
u
la
r/
w
el
l
P
o
in
t-
el
o
n
g
at
ed
H
ig
h
m
at
.,
lo
w
al
t.
(f
sp
,
Il
li
te
co
at
in
g
s,
k
ao
l*
),
p
o
ly
q
tz
H
il
ss
an
d
st
ei
n
(H
I)
Q
u
ar
tz
ar
en
it
e
9
5
/5
/0
M
o
d
.
S
y
n
t.
Q
tz
.
R
o
u
n
d
ed
/
m
o
d
.
C
o
n
ca
v
e-
co
n
v
ex
H
ig
h
m
at
.,
lo
w
al
te
ra
ti
o
n
;
p
o
ly
q
u
ar
tz
,
g
la
u
co
n
it
e
B
en
th
ei
m
(B
B
)
Q
u
ar
tz
ar
en
it
e
9
5
/3
/2
F
ew
sy
n
t.
Q
tz
/
k
ao
l.
,
il
li
te
S
u
b
ro
u
n
d
ed
/
w
el
l
C
o
n
ca
v
e-
co
n
v
ex
H
ig
h
m
at
.,
k
ao
li
n
it
e,
sw
el
la
b
le
cl
ay
m
in
er
al
s,
p
o
ly
q
tz
.
T
ri
as
si
c
U
p
p
er
K
eu
p
er
R
h
a¨t
V
el
p
k
e
(V
E
)
Q
u
ar
tz
ar
en
it
e
9
5
/5
/0
D
is
ti
n
ct
sy
n
t.
q
tz
.
S
u
b
ro
u
n
d
ed
/
v
er
y
w
el
l
P
o
in
t-
el
o
n
g
at
ed
H
ig
h
m
at
.,
p
o
re
fi
ll
in
g
k
ao
li
n
it
e,
al
te
ra
ti
o
n
o
f
fs
p
S
ee
b
er
g
en
(S
E
)
Q
u
ar
tz
ar
en
it
e
9
8
/2
/0
Q
tz
,
fe
w
ir
o
n
o
x
id
es
S
u
b
ro
u
n
d
ed
/
v
er
y
w
el
l
C
o
n
ca
v
e-
co
n
v
ex
V
er
y
h
ig
h
m
at
.,
v
er
y
lo
w
al
te
ra
ti
o
n
,
fe
w
k
ao
l.
R
o¨
h
n
b
er
g
(R
O¨
)
q
u
ar
tz
ar
en
it
e
9
5
/3
/2
K
ao
li
n
it
e,
il
li
te
S
u
b
ro
u
n
d
ed
/
v
er
y
w
el
l
P
o
in
t-
el
o
n
g
at
ed
H
ig
h
m
at
.,
lo
w
al
t.
(s
w
el
l.
cl
ay
,
p
o
re
fi
ll
.
k
ao
l.
)
H
o
ck
el
n
(H
O
)
Q
u
ar
tz
ar
en
it
e
9
8
/2
/0
Q
tz
,
cl
ay
S
u
b
ro
u
n
d
ed
/
v
er
y
w
el
l
E
lo
n
g
at
ed
H
ig
h
m
at
.
L
o
w
al
t.
S
ch
il
fs
a
n
ds
te
in
B
ar
k
h
au
se
n
(B
A
)
L
it
h
.
su
b
ar
k
o
se
7
5
/1
0
/
1
5
Q
tz
,
fe
w
ca
rb
o
n
at
e
S
u
b
an
g
.
/
w
el
l
C
o
n
ca
v
e-
co
n
v
ex
lo
w
al
te
ra
ti
o
n
;
ch
lo
ri
te
/v
o
lc
an
ic
cl
as
ts
S
ch
o¨
tm
ar
(S
C
O¨
)
L
it
h
.
su
b
ar
k
o
se
5
6
/2
0
/
2
4
Q
tz
,
fe
w
ca
rb
o
n
at
e,
h
em
at
it
e
S
u
b
an
g
.
/
m
o
d
.-
w
el
l
C
o
n
ca
v
e-
co
n
v
ex
L
o
w
al
te
ra
ti
o
n
;
ch
lo
ri
te
/v
o
lc
an
ic
cl
as
ts
H
in
d
fe
ld
en
(H
S
)
F
el
d
sp
.
L
it
h
ar
en
it
e
6
0
/1
0
/
3
0
L
it
tl
e
q
tz
S
u
b
an
g
.
/
w
el
l
P
o
in
t
L
o
w
m
at
.,
h
ig
h
al
t.
fs
p
&
cl
as
ts
,
il
l.
,
ch
l,
k
ao
l.
,
m
ix
ed
la
y
er
S
an
d
er
(S
A
)
F
el
d
sp
.
L
it
h
ar
en
it
e
6
5
/1
0
/
2
5
S
y
n
t.
q
tz
,
fs
p
,
cl
ay
S
u
b
an
g
.
/
m
o
d
.-
w
el
l
P
o
in
t-
el
o
n
g
at
ed
L
o
w
m
at
.,
h
ig
h
(a
lt
er
ed
fs
p
,
cl
ay
m
at
ri
x
,
m
ix
ed
la
y
er
*
,
ch
lo
ri
te
In
g
er
sl
eb
en
(I
S
)
L
it
h
ar
en
it
e
6
0
/1
0
/
3
0
C
la
y
le
y
m
at
ri
x
,
fe
w
sy
n
t.
q
tz
su
b
an
g
.
/
m
o
d
.-
w
el
l
p
o
in
t
Il
li
te
,
ch
lo
ri
te
,
k
ao
li
n
it
e
S
ch
le
er
ie
th
(S
L
)
L
it
h
ar
en
it
e
6
5
/5
/3
0
F
ew
sy
n
t.
q
tz
/
p
se
u
d
o
m
at
ri
x
S
u
b
an
g
./
m
o
d
.-
w
el
l
E
lo
n
g
at
ed
-
p
o
in
t
L
o
w
m
at
.,
h
ig
h
al
te
ra
ti
o
n
:
fs
p
,
cl
ay
m
at
ri
x
,
m
ic
a,
ch
lo
ri
te
L
o
w
er
B
u
n
ts
an
d
st
ei
n
S
o
ll
in
g
K
ar
ls
h
af
en
(K
A
)
S
u
b
li
th
ar
en
it
e
5
2
/1
7
/
2
7
S
y
n
t.
q
tz
,
cl
ay
&
fe
rr
it
ic
,
C
a
A
n
g
u
la
r/
m
o
d
.-
w
el
l
E
lo
n
g
at
ed
-
co
n
c.
M
o
d
.
m
at
.,
m
o
d
.
al
te
ra
ti
o
n
,
sm
ec
ti
te
*
,
k
ao
li
n
it
e*
A
rh
o
lz
en
(A
H
)
S
u
b
li
th
ar
en
it
e
8
0
/1
0
/
1
0
S
y
n
t.
q
tz
,
cl
ay
&
fe
rr
it
ic
,
C
a
A
n
g
u
la
r/
m
o
d
.-
w
el
l
E
lo
n
g
at
ed
M
o
d
er
at
e
m
at
.,
m
o
d
er
at
e
al
te
ra
ti
o
n
;
se
d
.
cl
as
ts
L
o
b
ac
h
(L
O
)
S
u
b
li
th
ar
en
it
e
8
0
/1
0
/
1
0
S
y
n
t.
q
tz
,
cl
ay
A
n
g
u
la
r/
m
o
d
.-
w
el
l
E
lo
n
g
at
ed
M
o
d
er
at
e
m
at
.,
m
o
d
er
at
e
al
te
ra
ti
o
n
;
se
d
.
cl
as
ts
U
d
er
(U
D
)
Q
u
ar
tz
ar
en
it
e
9
0
/5
/5
Q
tz
,
cl
ay
fe
rr
it
ic
A
n
g
u
la
r/
m
o
d
.-
w
el
l
P
o
in
t
(e
lo
n
g
at
ed
)
m
o
d
er
at
e
m
at
.,
lo
w
al
t.
,
se
d
.
cl
as
ts
Environ Earth Sci (2013) 69:1299–1332 1305
123
formed lithic subarkoses and sublitharenites occur.
According to Strack and Stapf (1980), they are fluviatile
deposits from flood plains and channel facies.
Triassic sandstones of the Buntsandstein are terrigeneous
deposits of predominantly sublitharenitic composition.
Sandstones of the Middle Buntsandstein commonly show
quartz arenitic to sublitharenitic composition. The deposi-
tional environment of the Bunter sediments changes mark-
edly through time. In the Lower Bunter playa facies, braided
river systems formed at the margins (Paul 1982), which are
characterized by fining upward cycles. The sandstones of the
Middle Buntsandstein are characterized by prograding flu-
vial sandstones with a common channel facies (Karlshafen,
Arholzen, Uder and Lobach). Sandstones of the Upper
Bunter are seldom used as building stones because they
mainly consist of clayey and silty deposits.
Sandstones of Upper Triassic (Keuper) age are charac-
terized by marked changes of the depositional environment
through time. They also document a great variability in
their physical properties. Considerable natural stone
deposits belong to the Lettenkohlenkeuper, Schilfsandstein
and Rha¨t Formations (Lower, Middle and Upper Keuper).
Sandstones deposited during the Schilfsandstein and
Lettenkohlenkeuper are mainly litharenites and sublit-
harenites, commonly with highly altered unstable rock
fragments. Sandstones of the Lettenkohlenkeuper (Inger-
sleben and Schleerieth sandstones) were deposited in a
brackish marine delta environment with local moors. Local
coal layers also occur with the same name designation
(Beutler and Schubert 1987; Paul et al. 2008). Sediments of
the Schilfsandstein (Sander, Hindfelden, Scho¨tmarer and
Barkhausen sandstones) were deposited in a wide delta
facies covering a flat shelf. They formed meandering delta
arms filled by up to 50 m of fine sand (Wurster 1964). Rha¨t
sandstones are characterized by quartz arenites (the Vel-
pke, Hockeln, Seebergen and Ro¨hnberg sandstones). They
were deposited under changing depositional environments
from marine to lacustrine-terrestrial. The sediments are
predominantly composed of alternating fine sands and
mudstones (Seidel 2003). Very coarse channel fills also
occur (Koch et al. 2003; Sobott and Koch 2009) in the
Franconian Rha¨t deposits.
Cretaceous sandstones (Cotta, Ru¨then, Hilssandstein,
Obernkirchen and Bad Bentheim) were deposited in the
coastal rim area formed by a transgression from the north.
They mainly consist of quartz arenites, in which the marine
environment is partially well documented by the occur-
rence of glauconite (greensands).
Petrographical analyses—composition and fabric
Compositionally, the sandstones investigated can generally
be defined as quartz arenites, sublitharenites, litharenitesT
a
b
le
1
co
n
ti
n
u
ed
S
am
p
le
C
la
ss
ifi
ca
ti
o
n
M
cB
ri
d
e
(1
9
6
3
)
Q
tz
/F
sp
/
L
it
h
o
C
em
en
t/
m
at
ri
x
R
o
u
n
d
n
es
s/
so
rt
in
g
G
ra
in
co
n
ta
ct
M
at
u
ri
ty
&
d
eg
re
e
o
f
al
te
ra
ti
o
n
,
cl
ay
m
in
er
al
s,
ro
ck
fr
ag
m
en
ts
P
er
m
ia
n
L
o
w
er
R
o
tl
ie
g
en
d
B
eb
er
ta
l
(B
E
)
S
u
b
li
th
ar
en
it
e
7
8
/8
/1
4
C
la
y
fe
rr
it
ic
,
fe
w
sy
n
t
q
tz
&
C
a
S
u
b
ro
u
n
d
ed
/
m
o
d
.
P
o
in
t-
el
o
n
g
at
ed
M
o
d
er
at
el
y
w
el
l
m
at
u
ri
ty
,
lo
w
al
te
ra
ti
o
n
,
sm
ec
ti
te
*
T
am
b
ac
h
(T
A
)
S
u
b
li
th
ar
en
it
e
9
0
/2
/8
S
y
n
t.
q
tz
,
ca
lc
.,
fe
rr
it
ic
S
u
b
ro
u
n
d
ed
/
m
o
d
.
P
o
in
t-
el
o
n
g
at
ed
M
o
d
er
at
el
y
w
el
l
m
at
u
ri
ty
,
lo
w
al
te
ra
ti
o
n
,
ch
lo
ri
te
*
C
ar
b
o
n
if
er
o
u
s
Ib
b
en
b
u¨
re
n
(I
B
)
S
u
b
li
th
ar
en
it
e
8
0
/2
/1
8
Q
tz
,
cl
ay
fe
rr
it
ic
S
u
b
an
g
u
la
r/
w
el
l
C
o
n
ca
v
e-
co
n
v
ex
M
o
d
er
at
e
m
at
.,
fe
w
al
te
ra
ti
o
n
(f
sp
.)
,
k
ao
l
&
m
o
n
tm
o
ri
ll
o
n
it
e*
*
A
lt
al
te
ra
ti
o
n
,
fs
p
fe
ld
sp
ar
,
ka
o
l
k
ao
li
n
it
e,
m
a
t
m
at
u
re
,
m
o
d
m
o
d
er
at
e,
p
o
ly
q
tz
p
o
ly
q
u
ar
tz
,
se
d
se
d
im
en
ta
ry
,
sw
el
l
sw
el
la
b
le
,
sy
n
.
q
tz
sy
n
ta
x
ia
l
q
u
ar
tz
o
v
er
g
ro
w
th
(s
il
ic
a
ce
m
en
t)
*
B
ar
te
ls
en
(2
0
0
8
)/
R
u
ed
ri
ch
et
al
.
(2
0
1
1
)
*
*
S
ei
d
el
(2
0
0
4
)/
R
u
ed
ri
ch
et
al
.
(2
0
0
5
)
1306 Environ Earth Sci (2013) 69:1299–1332
123
and lithic subarkoses according to McBride (1963, Fig. 3).
Arkoses and lithic arkoses are very rare in this data set. The
22 sandstones which were analysed in detail comprise a
wide range of different sandstone types and are represen-
tative for the entire dataset compiled from the literature
(*300 samples) in terms of composition, depositional
settings and age.
The Ibbenbu¨ren sandstone (Fig. 3; Table 1) of Carbon-
iferous age shows a sublitharenitic composition and is less
mature due to its clayey matrix, subangular–subrounded
grain shape and well to moderate sorting. The grain contact
is predominantly sutured but also concave–convex
(Fig. 5a) and visible pore spaces are only moderately rec-
ognizable. Cementation occurs as syntaxial quartz over-
growths, whereby the original grain shape is marked by a
thin rim of impurities. The Permian Tambach and Bebertal
sandstones are both characterized by a sublitharenitic
composition. The Bebertal aeolian sandstone only shows
moderate sorting (Figs. 4a, 6) with a bimodal grain size,
whereby the larger grains are rounded and the smaller
grains rather subrounded. Grain contacts are mostly elon-
gated–concave–convex. Syntaxial quartz overgrowths
(Fig. 5a, c) and calcite occur as cement. Locally, the grains
are surrounded by early infiltrated clay minerals (illite).
Furthermore, X-ray diffraction indicates smectite as the
clay mineral found in these sandstones. Cation exchange
capacity (CEC) for the whole rock fraction amounts to
1.0 meq/100 mg (Bartelsen 2008). The Tambach sandstone
is characterized by well sorting and subrounded–subangu-
lar grain shapes, concave–convex grain contacts and very
well-pronounced syntaxial quartz overgrowths. Further-
more, calcite occurs as a pore-filling cement and feldspar
as a grain overgrowth. In addition, detrital grains are sur-
rounded by finely dispersed iron oxides. Partially altered
lithoclasts show replacement by chlorite, which was also
determined by XRD analyses. The cation exchange
capacity is with 1.1 meq/100 mg marginally higher than
for the Bebertal sandstone (Bartelsen 2008).
The fluviatile Uder, Lobach, Arholzen and Karlshafen
sandstones belong to the Lower Triassic of Buntsandstein
Fig. 3 Classification of sandstones investigated after McBride (1963)
and Hayes (1984). b IB Ibbenbu¨ren sandstone, Ta Tambach
sandstone, Be Bebertal sandstone, Ve Velpke sandstone, CO Cotta
sandstone, RU¨ Ru¨then sandstone, BB Bad Bentheim sandstone, Hi
Hilssandstein, Ob Obernkirchen sandstone; c LO Lobach sandstone,
AH Arholzen sandstone, KA Karlshafen sandstone, UD Uder sand-
stone; d SL Schleerieth sandstone, IS Ingersleben sandstone, HS
Hindfelden sandstone, SA Sander sandstone, SCO¨ Scho¨tmar sand-
stone, BAR Barkhausen sandstone, RO¨ Ro¨hnberg sandstone, SE
Seebergen sandstone, HO Hockeln sandstone
Environ Earth Sci (2013) 69:1299–1332 1307
123
(Table 1). They are sublitharenites in composition with the
exception of the Uder sandstone, which is defined as a
quartz arenite. While the samples of Lobach, Arholzen and
Karlshafen (Weser sandstones) all show distinct alteration
of feldspars and lithoclasts, the Uder sandstone shows only
little alteration. The three Wesersandstones are character-
ized by a subangular grain shape, predominantly concave–
convex grain contact, syntaxial quartz overgrowth, calcite
cement, locally feldspar cement and often a haematite
overprinting of the grain surfaces (Fig. 4b). However, the
amount of cement and also the degree of compaction var-
ies. Most of the (quartz-) cementation occurs in the Karl-
shafen sandstone, followed by Arholzen and finally
Lobach. Furthermore, the Karlshafen sandstone shows the
strongest compaction. The grains of the Uder sandstone are
subrounded, grain contacts pointed-elongated and little
stabilizing cement is present; only a clay matrix is recog-
nizable. This sandstone shows little compaction since the
grain contacts are predominantly pointed. The sorting and
grain size of the Bunter sandstones exhibits distinct vari-
ations, whereas the Lobach, Arholzen and Karlshafen
sandstones (Wesersandstones) show a moderate to well
sorting. The Uder sandstone (Thuringia) is well sorted
(Fig. 6). Weser sandstones can be classified as fine sand,
whereas the Karlshafen sandstone even shows a fine-silty
grain size. The Uder sandstone exhibits a fine–medium
grain size. In general, the Uder sandstone is composition-
ally and texturally more mature than the distinct immature
Weser sandstones. The cation exchange capacity for the
Karlshafen sandstone equals 2.0, and XRD analyses indi-
cate the presence of chlorite, smectite and kaolinite in the
clay fractions \2 lm (Bartelsen 2008).
Both the Schleerieth and Ingersleben sandstones belong
to the Lower Keuper (Table 1). They are characterized by a
litharenitic composition. The Schleerieth sandstone exhib-
its little syntaxial quartz overgrowths, a clayey matrix,
Fig. 4 Representative
sandstone types: a Aeolian
sandstone (dune deposit;
Rotliegend, Bebertal sandstone)
revealing a bimodal grain size
distribution (x-pol.).
b Sublitharenite with quartzitic
and calcareous cementation as
well as a haematitic
overprinting, which is
characteristic for fluviatile
sandstones belonging to the
Lower Triassic (Karlshafen
sandstone; x-pol.). c Highly
immature sublitharenite–
litharenite with altered
lithoclasts, feldspars as well as
pseudomatrix (Hindfelden
sandstone; II-pol.). d Quartz
arenite (Upper Triassic/Ra¨t,
Seebergen) showing clean grain
contacts, which allows pressure
solution leading to concave–
convex to sutured grain contacts
(x-pol.). e Quartz arenite with
early infiltration of illite
forming clay boundaries
between grains (Ro¨hnberg
sandstone; x-pol.). f Marine
sandstone with glauconite
grains (Cretaceous, Ru¨then;
x-pol.)
1308 Environ Earth Sci (2013) 69:1299–1332
123
predominantly point-elongated grain contacts and sub-
rounded grains. Feldspar grains are sericitized and the
mainly volcanogenic rock fragments are distinctly altered.
Chloritization of volcanogenic rock fragments is also
observable. The grain size is coarse silty to fine sandy and
the sorting well developed (Fig. 6). XRD analyses show
chlorite and illite within the sandstone and its CEC
amounts to 2.1 meq/100 g (Bartelsen 2008). Thus, this
rock can be defined as very immature sandstone. The
Ingersleben sandstone can also be defined as highly
immature and is characterized by little compaction and
cementation; however, the rock shows a distinct visible
pore space. The volcanogenic rock fragments present are
highly altered to chlorite and the feldspars show alteration
by sericitization. Often, loose fragments of pseudomatrix
between the few stable grains can be observed, and rarely a
grain contact between stable components is visible.
Locally, relicts of feldspar cement are recognizable, which
are also now replaced. The grain contact is mainly pointed
or elongated and the grain shape subangular and locally
angular. Furthermore, at grain surfaces, chlorite and pore-
lining illite are observable. XRD analyses indicate the
presence of chlorite, kaolinite and illite (Fig. 7) and its
CEC amounts to 1.4 meq/100 g.
Sandstones of the Middle Keuper (Table 1) are pre-
dominantly compositionally immature, since the Sander and
Hindfelden sandstones are feldspathic litharenites and the
Barkhausen and Scho¨tmar sandstones lithic subarkoses.
Fig. 5 Characteristic features for grain contacts and cementation.
a Syntaxial quartz overgrowth as well as sutured point contact
between quartz grains (II-pol.). b Clay coatings of needle-shaped illite
lining the pores (x-pol.). c Quartz overgrowth, concave–convex grain
contact as well as dark brown ferrous oxides present as rims and
interparticle fillings (II-pol.). d Distinct clay coatings of chlorite at
altered grain surfaces (II-pol.). e Squeezed pseudomatrix between
altered detrital grains (II-pol.) and f distinct quartz overgrowths under
cathodoluminescence
Environ Earth Sci (2013) 69:1299–1332 1309
123
The Barkhausen, Scho¨tmar and Sander sandstones contain
quartz cement, locally feldspar cement and a clayey matrix,
whereas the Hindfelden sandstone shows very little quartz
cement and a distinct pseudomatrix (Figs. 4c, 5e). The
grain surfaces of the Barkhausen, Scho¨tmar and Sander
sandstones are overprinted with haematite, while the grain
surfaces of the Hindfelden sandstone are overprinted by
thick clay–chlorite rims (Fig. 5d). Alteration of feldspar
and unstable rock fragments are distinctly developed in the
latter sample and rarely a grain contact between stable
quartz grains can be observed. Instead of this, quartz grains
commonly are floating in the pseudomatrix. In contrast to
the Barkhausen and Scho¨tmar sandstones that show con-
cave–convex grain contacts, the Sander sandstone exhibits
elongated grain contacts. All Middle Keuper sandstones
exhibit a well to moderate sorting and subrounded–sub-
angular grain shape. The grain size of the Barkhausen
sandstone is about 100 lm, the Scho¨tmar sandstone ranges
100–200 lm, the Hindfelden standstone shows sizes of
80–100 lm and the sandstone Sander has a size of around
175 lm. XRD analyses show a large chlorite and illite
content and a small amount of kaolinite in the Hindfelden
sandstone (Fig. 7), where its CEC amounts to 4.6 meq/
100 g. The CEC amounts to Sander sandstone 3.4 meq/100 g
and swellable mixed layer minerals and kaolinite are
detected by XRD analyses.
Seebergen, Ro¨hnberg, Hockeln and Velpke sandstones
of the Upper Keuper (Rha¨t) are all quartz arenites in
composition, well to very well sorted, rounded to sub-
rounded and classified as fine, mature sands. They can be
Fig. 6 Sorting of sandstones investigated: a Permian and Carboniferous sandstones, b Buntsandstein and Lower Keuper, c Upper Keuper and
d Cretaceous. For sample abbreviations, see Table 1
1310 Environ Earth Sci (2013) 69:1299–1332
123
distinguished by their different degrees of compaction,
early infiltrated clay minerals and cementation. Seebergen
sandstones are moderate–strongly compacted, show con-
cave–convex grain contacts, clean grain surfaces (Figs. 4d,
5a, c) and distinct syntaxial quartz overgrowths. In con-
trast, the Ro¨hnberg sandstone is characterized by little
compaction and quartz cementation, but also shows a dis-
tinct early infiltrated grain lining with illite and kaolinite
pore fillings (Fig. 4e). The Hockeln sandstone is mainly
characterized by quartz cementation, low clayey matrix
contents, little infiltrated early clay minerals and sericitized
feldspars. Sandstone Velpke is characterized by distinct
amounts of syntaxial quartz overgrowth, iron oxides and
pore-filling kaolinite. The feldspars are altered to kaolinite.
Grain contact of the Ro¨hnberg sandstone predominantly is
punctiform, whereas the Seebergen exhibits concave–con-
vex, the Velpke pointed–elongated and the Hockeln elon-
gated–concave–convex grain contacts. The average grain
size of the Seebergen ranges from 150 to 200 lm, the
Ro¨hnberg about 100 lm and of sandstone Velpke and
Hockeln around 125 lm. Results of XRD analyses for the
Seebergen indicate traces of kaolinite, but no other clay
minerals (Fig. 7) and with a very low CEC of 0.2 meq/
100 g. In contrast, the Ro¨hnberg sandstone shows high
kaolinite contents and also swellable clay minerals (Fig. 7),
where the CEC equals 1.5 meq/100 g.
Cretaceous sandstones investigated in this study
(Table 1) are the Cotta, Ru¨then, Bad Bentheim, Hilssand-
stein and Obernkirchen marine and mature sandstones,
which all show a quartz arenitic composition. Polycrys-
talline quartz grains often occur and in the case of the
Ru¨then, Cotta and Hilssandstein sandstones glauconite
grains are present (Fig. 4f). After Tucker (2001), sandstone
Ru¨then has been defined as hybrid sandstone, since here the
glauconite occurs to a significant proportion. The few
feldspars that occur are only slightly altered by sericitiza-
tion. With the exception of the Obernkirchen sandstone,
whose grains are subangular, the sandstones are sub-
rounded to round. Sandstone Obernkirchen also shows very
high amount of quartz cementation, whereas at the other
sandstones syntaxial quartz cementation is much lower. In
addition, the Cotta sandstone contains a clayey matrix and
distinct altered feldspars and therefore is more immature.
Ru¨then, Bad Bentheim and Obernkirchen sandstones
exhibit pore-lining illite and pore-filling kaolinite (Fig. 5b).
The grain contact of the Bad Bentheim sandstone is gen-
erally pointed, but also concave–convex contact occurs
(Fig. 5a). In contrast, the Ru¨then and Obernkirchen sand-
stones mainly show concave–convex grain contacts, and
the Cotta and Hilssandstein sandstones predominantly
reveal elongated to concave–convex grain contacts. Lowest
grain sizes are found in the Obernkirchen sandstone with
an average of around 80 lm, whereas the largest occur in
the Bad Bentheim sandstone with an average of around
250 lm. The sorting of these sandstones is well pro-
nounced. Results from the XRD analyses confirm the
presence of high kaolinite contents and swellable clay
minerals in the Bad Bentheim sandstone, but the CEC only
shows a very low value of 0.1 meq/100 g. Furthermore, the
sqrt illustration overemphasises the very low intensities of
illite (Fig. 7b). In the Bad Bentheim sandstone, illite occurs
as mixed phase along with swellable clay minerals.
Fig. 7 Mineral content of clay fractions (textural preparates) from
the Seebergen, Ingersleben, Hindfelden, Bad Bentheim and Ro¨hnberg
sandstones measured at textural specimen. Blue line/red: ethylene
glycolate; black line: air dried. Red rectangle marks the area with the
presence of swellable clay minerals e.g. in sandstone Bad Bentheim
Environ Earth Sci (2013) 69:1299–1332 1311
123
Petrophysical parameters and weathering behaviour
Univariate and bivariate distribution
Whisker plots document the distribution of all samples of
one rock property (Fig. 8). Most of the data exhibit a
symmetrical distribution, with the exception of the coeffi-
cient of water absorption and the compressive strength.
The mean and the median of the matrix density is
2.66 g/cm3 (third quartile 2.64 g/cm3 respectively 2.68 g/
cm3), a value that corresponds to the high amount of quartz
and feldspar. Deviations from these data are caused by the
presence of non-quartzitic rock fragments and by the
occurrence of authigenetic minerals such as chlorite, cal-
cite, haematite or glauconite that usually occur in marine
sandstones. According to this, outliers observable probably
can be explained by data of immature sandstones. The bulk
density is characterized by a wide scattering (1.78–2.45 g/
cm3) reflecting the highly variable porosity of sandstones
(1–33 %; mean 16.8 %; first quartile 12 %, third quartile
22 %), with an average value of 2.14 g/cm3. Both the
matrix and bulk density show a very low variance
(Table 2). Porosity data are characterized by a normal
distribution with a median and mean of 17.45 %. Fifty
percent of the data range between 12 and 22 % porosity
and outliers do not occur. The water absorption is char-
acterized by a smaller range of quartiles and whiskers, but
also shows extreme outliers. The median water absorption
under atmospheric conditions (5.4 wt%) is markedly lower
than the water absorption under vacuum (8.5 wt%). Fur-
thermore, some outliers in both directions are observable.
The coefficient of water absorption (w value) is charac-
terized by a mean of 6.25 kg/ms-2 and a median of 4.3 kg/
ms-2, by 2.2 kg/ms-2 for the first quartile and 7.5 kg/ms-2
for the third quartile. Many outliers are also observable
here, but only upwards. The position of the median defines
a skewed distribution. The Bad Bentheim sandstone shows
the highest coefficient of water absorption with a value of
about 40 kg/ms-2. The saturation index varies from 0.4 to
0.99 (mean 0.67, median 0.68) and only shows one outlier.
The compressive strength of all sandstones shows a median
of 43 MPa and a mean of 51.68 MPa (first quartile
34 MPa, third quartile 63 MPa). The lowest compressive
strength is 12 MPa and the highest one 107 MPa. How-
ever, some sandstones reach values of up to 160–180 MPa,
such as the Permian and Middle Keuper sandstones of
Tambach and Barkhausen.
Porosity and compressive strength show marked differ-
ences and interrelations between each other, taking into
account different lithotypes and their stratigraphic position
as documented in Fig. 9 for four selected characteristic
sandstones. Poorly altered and strongly compacted sublit-
harenites such as the Karlshafen and Arholzen Weser
sandstones (Fig. 9b; Tables 1, 3) show the lowest porosity.
Pure quartz arenites (mean and median porosity: 15 %)
include sandstones with distinct quartz overgrowths and no
early infiltration by clay. Sandstones of the Upper Keuper
(SE) and of Cretaceous age (OB) commonly fall into this
Fig. 8 Whisker plots of petrophysical properties from: matrix
density, bulk density and effective porosity [n = 305; Chitsazian
(1985), David (2006), Grimm (1990), Holzwarth (1996), Kirsten
(2009), Massallam (1973), Mosch and Siegesmund (2007) and own
samples], water uptake under atmospheric and vacuum conditions,
capillary water uptake [n = 214; David (2006), Grimm (1990),
Kirsten (2009) and own samples], s value of 286 as well as
compressive strength [n = 246; David (2006), Kirsten (2009), Mosch
and Siegesmund (2007), Holzwarth (1996) and own samples]. The
red box indicates the range at which 50 % of the mean data are found
and the black bar within the box the median. The red dots represent
outliers, which are outside the interquartile ranges
1312 Environ Earth Sci (2013) 69:1299–1332
123
group. Weakly compacted and weakly cemented quartz
arenites (median 24 %) and strongly altered sublithare-
nites–litharenites (median porosity 25 %) show the highest
porosity data. The latter group displays the most distinct
scattering and is predominantly represented by sandstones
of Lower and Middle Keuper (HS and IS; Tables 1, 3).
Considering the stratigraphic position of the sandstones
analysed, the following relationships can be delineated.
With increasing stratigraphic age, the porosity generally
decreases, indicating that porosity is a function of time and
burial depth (e.g. Wilson 1994, Koch and Sobott 2008).
Sandstones of the Buntsandstein show a median porosity of
17 %. Those of the Lower and Middle Keuper exhibit the
highest porosities (a mean of 24 %). Similar porosities
exist in the sandstones of the Upper Keuper and the Cre-
taceous, with a mean porosity of 17.5 %. Therefore,
sandstones of the Upper Keuper and the Cretaceous do not
follow the time/burial trend as shown by elder sandstones.
This is probably due to the fact that at low burial, the
primary parameters of depositional environment and of
Fig. 9 Whisker plots of a porosity ordered according to lithotype and
to stratigraphic age [n = 258; David (2006), Fitzner (1970), Kirsten
(2009), Holzwarth (1996) and this study] and b compressive strength
ordered according to lithotype and stratigraphic age. [n = 258;
Chitsazian (1985), David (2006), Fitzner (1970), Kirsten (2009),
Holzwarth (1996) and this study]
Table 2 Mean and variances of data distributions of matrix and bulk density, porosity, water absorption under atmospheric and vacuum
conditions, coefficient of water absorption, saturation index and compressive strength
Matrix density
(g/cm3)
Bulk density
(g/cm3)
Porosity
vol. (%)
Water abs.
(atm.) vol. (%)
Water abs.
(vac.) (%)
Coefficient of
water absorption
Saturation
index (-)
Compressive
strength (MPa)
Mean 2.66 2.14 17.45 7.41 8.47 6.25 0.68 51.68
Variance 0.01 0.02 34.67 4.25 8.29 39.02 0.01 767.96
Data from David (2006), Fitzner (1990), Grimm (1990), Kirsten (2009), Holzwarth (1996), Mosch and Siegesmund (2007) and own sandstones
Environ Earth Sci (2013) 69:1299–1332 1313
123
eogenetic cementation as well as early formed secondary
porosity always have a major effect. Moreover, they are not
completely overprinted and hidden by deeper burial alter-
ation and homogenization.
Also, the distribution of compressive strength was
grouped according to the lithotypes and stratigraphic
position (Fig. 9b). The median for strong compacted and/or
cemented quartz arenites amounts to 87 MPa, whereas
weak compacted quartz arenites exhibit with 56 MPa a
distinct lower compressive strength. Sublitharenites, which
contain 5–25 % unstable lithic rock fragments, are strongly
compacted and cemented and have a median compressive
strength of 103 MPa. Weakly compacted and cemented,
altered sublitharenites–litharenites exhibit the lowest
compressive strength with 37 MPa as median. The com-
pressive strength dependent on stratigraphy shows for
Permian and Lower Triassic (Buntsandstein) sandstones
the highest values, whereas in sandstones belonging to
Middle and Lower Keuper (Upper Triassic) the lowest
compressive strength is observable (median: 37 MPa).
Table 3 Petrophysical characteristics of all 22 sandstones analyzed. These data together with numerous data from the literature is the database
for the interrelationships of selected petrophysical parameters and weathering resistance (see Fig. 18/19)
Sample Porosity
[%]
W-value
[kg/m*s2]
Compr.
strength
[MPa]
Pore radii
distribution
Hygric
dilatation
[mm/m]
Group of
salt resistance
Cretaceous Upper Cotta 25.20 6.10 45.00** Bimodal 0.01 C**
Ru¨then 25.60 20.40 40.00** Bimodal 0.18 C**
Lower Obernkirchen 17.50 2.00 100.00 Unimodal
equable
0.11 A
Hilssandstein 20.80 24.90 60.00** Unimodal
equable
0.02 C**
Bentheim 25.60 40.70 95.00 Unimodal
equable
0.02 A
Upper Triassic Rha¨t Velpke 24.70 33.40 50.00** Unimodal
equable
0.03 B**
Seebergen 15.80 1.80 108.91 Unimodal
unequal
0.01 A
Ro¨hnberg 21.87 11.85 32.83 Unimodal
equal
0.15 C
Hockeln 15.40 3.40 65.00** Bimodal 0.19 C**
Schilfsandstein Barkhausen 3.80 0.40 170.00* Bimodal 0.93 A
Scho¨tmar 10.30 0.90 80.00* Bimodal 1.11 B**
Hindleben 22.07 18.05 16.10 Bimodal 5.21 C
Sander 17.00 20.05 54.71 Bimodal 1.25 C
Letten-kohlenkeuper Ingersleben 27.71 31.03 23.10 Unimodal
equable
2.80 C
Schleerith 15.26 2.01 55.00 Bimodal 0.98 C
Lower Triassic Middle Buntsandstein Karlshafen 7.90 0.50 119.14 Unimodal
unequal
1.35 A
Arholzen 10.68 1.20 79.43 Unimodal
unequal
1.39 A
Lobach 11.94 1.35 94.00* Unimodal
unequal
0.60 B
Uder 23.26 19.59 48.00* Unimodal
equable
0.58 B
Permian Lower Rotliegend Bebertal 9.68 0.93 68.00 Unimodal
unequal
0.54 B
Tambach 8.28 0.58 160.00* Bimodal 0.57 A
unimodal
Carboniferous Ibbenbu¨ren 12.50 1.00 120.00** Unimodal
unequal
0.53 A**
* Bartelsen (2008)/Ruedrich et al. (2011)
** Seidel (2004)/Ruedrich et al. (2005)
1314 Environ Earth Sci (2013) 69:1299–1332
123
Three different types of sandstones can be recognized
based on the pore radii distribution (I—bimodal, II—uni-
modal unequal and III—unimodal equal). Predominantly,
litharenites of Upper Triassic age (Lower and Middle
Keuper, HS, SA, SL; Tables 1, 3) with very low maturity
show a bimodal pore radii distribution (Fig. 10I). Micro-
sized pores are caused by the alteration of less stable
components and/or the presence of a pseudomatrix. The
second maximum of capillary pores reflects relics of pri-
mary intergranular porosity, which was closed by different
degrees due to cementation and compaction. Unimodal
unequal (Fig. 10II) distributions of pore radii can be pre-
dominantly observed in sublitharenites, but they can also
occur in quartz arenites. In sublitharenites, this pore radii
distribution can be attributed to moderate to well sorting,
weak alteration of unstable grains and a moderate degree of
cementation and compaction as documented by the Weser
sandstones of Lower Triassic age (Buntsandstein, KA, AH,
UD). Quartz arenites, which show such a distribution of
pore radii, are generally well sorted and exhibit a more
immature character due to the infiltration of clay minerals
(Cretaceous and Upper Triassic sandstones; Upper Keuper,
SE). Unimodal equal distributions of pore radii (Fig. 10III)
are mainly found in well-sorted highly mature quartz are-
nites, in which no infiltration of clay minerals occurred as
documented by the Cretaceous sandstones (VE, RU¨, HI,
OBI). In general, an increase of maturity is correlated with
the pore radii distribution from bimodal to unimodal
unequal to unimodal equal.
No correlation seems to exist between grain size and
compressive strength or between grain size and porosity
(Fig. 11). According to Schießl and Alfes (1990), sandstones
with smaller grain size should exhibit higher strength, since
the amount of grain contacts per volume should be higher in
finer grained sandstones. Indeed, Schießl and Alfes (1990)
take into account the number of grain contacts. However, the
kind of grain contact stays unconsidered. Thereby, a quartz–
quartz grain contact should behave different from a quartz–
lithoclast, or quartz–feldspar grain contact. In consequence,
independent of grain size, the lithology (content of
mechanical stable/unstable grains) should have higher
impact on strength than grain size. Mu¨ller et al. (2002)
observed a linear relationship between porosity and grain
size, whereby with increasing grain size a decrease of
porosity should be observable. Based on the large data set in
this study, this cannot be confirmed. This investigation
assumes that modification of pore space during diagenesis
and initial sorting mainly rules the final porosity, whereby
the grain size has little effect.
The bivariate analysis of correlations between petro-
physical properties commonly reflects a linear or expo-
nential relationship (Fig. 12). In sandstones with similar
matrix density as the siliciclastic detrital grains, porosity
correlates to bulk density (correlation coefficient =
0.4114). High porosity is related to low bulk density and
vice versa (Fig. 12a). Newly formed mineral phases and/or
detrital minerals with different density (e.g. glauconite,
calcite, dolomite etc.) will result in a negative correlation.
The average value for all sandstones amounts to 15 %
porosity at a bulk density of 2.25 g/cm3. Compressive
strength decreases exponentially with an increasing coeffi-
cient of water absorption (Fig. 12b; correlation coeffi-
cient = 0.3497). Here, the average value of the probability
ellipsoid for all the sandstones investigated amounts to
70 MPa at 5 kg/ms-2. The water absorption, dependent on
porosity, increases linearly with a correlation coefficient of
R = 0.46 under vacuum and R = 0.37 under atmospheric
pressure (Fig. 12c). The ultrasonic velocity shows a linear
correlation (R = 0.4712) to porosity (Fig. 12d). High velocity
correlates with low porosity. The average value of 80 %
probability ellipsoid amounts to 2,700 m/s and a porosity of
13.5 %. Also, a linear relationship between ultrasonic velocity
and compressive strength is observable. Here, the velocity
increases with increasing compressive strength (Fig. 10e;
R = 0.6581). The average value of the ellipsoid equals
2,700 m/s and 50 MPa. The relationship between compres-
sive strength, flexural strength (R = 0.6074) and porosity is
illustrated in Fig. 12f: in general, both strengths decrease with
increasing porosity. Therefore, the gradient between porosity
and compressive strength is higher than the flexural strength.
Furthermore, a positive linear relationship between com-
pressive strength and flexural strength (Fig. 12g; R = 0.6074)
and between compressive strength and bulk density
(R = 0.3993) is obvious (Fig. 12h).
The regression analysis of basic parameters shows
besides the approximate linear relationships, a partially
distinct scattering. This distribution and interrelationship of
parameters can be explained in greater detail, when also
considering characteristics of petrographic composition
and fabric. These interrelationships between strength and
porosity to parameters such as pore radii distribution, grain
contact, degree of alteration (i.e. maturity) and content of
unstable lithic fragments is determined with confidence
regions of 80 % probability (Fig. 13). Sandstones with a
unimodal equal pore radii distribution in direction show a
higher compressive strength than sandstones with unimodal
unequal pore radii distributions (Fig. 13a). In contrast,
sandstones with a bimodal pore radii distribution show a
larger scattering of data as indicated by the ellipsoid with
80 % probability covering a wider range.
The interrelationships between compressive strength,
porosity and type of grain contacts are well documented
(Fig. 13b). Sandstones with low compressive strength and
high porosity mainly exhibit pointed and elongated grain
contacts. In contrast, sandstones with sutured and concave–
convex contacts exhibit higher compressive strengths.
Environ Earth Sci (2013) 69:1299–1332 1315
123
Although the average data of the compressive strength for
concave–convex grain contacts in the sandstones is lower
than the data for sutured grain contacts, outliers of sand-
stones with concave–convex grain contacts show strengths
with up to 170 MPa. Therefore, these two types of grain
contacts (concave–convex and sutured) can probably be
both considered equivalent and stabilizing. The interrela-
tionship between compressive strength and degree of
alteration to maturity is documented in Fig. 13c. Sublit-
harenites and litharenites with highly altered feldspars and
unstable lithic fragments and high amounts of chlorite in
pseudomatrix exhibit a very low compressive strength and
a high porosity. Quartz arenites, which show a low degree
of compaction and cementation, also display low com-
pressive strengths and high porosities. Besides the pore
radii distribution, type of grain contacts and the degree of
Fig. 10 Schematic documentation of pore radii distribution, devel-
opment of maturity and amount of micropores: (I) bimodal with two
maxima, (II) unimodal unequal and (III) unimodal equable. The bars
indicate the ratio of micropores to capillary pores. The vertical red
lines in the diagrams indicate the boundary between micro- and
capillary pores and the dashed line the average pore radius
Fig. 11 Diagrams showing the
relationship between porosity
and grain size, and between
compressive strength and grain
size dependent on sandstone
type. (Fitzner 1970, n = 22;
David 2006, n = 56; Kirsten
2009, n = 45; Grimm 1990,
n = 72, this study = 22)
1316 Environ Earth Sci (2013) 69:1299–1332
123
alteration, the compressive strength is also related to the
content of unstable lithic fragments. Generally, a decrease
of the compressive strength with increasing amounts of
unstable lithic fragments is observable as shown by sand-
stones with up to 5, 15 and 30 % unstable lithic fragments.
Furthermore, the interrelationship of these three groups to
porosity is indicated by the inclination of the three corre-
lation lines (Fig. 13d). The main reason for this interrela-
tionship is the content of unstable lithic fragments such as
volcanic and clayey clasts, which commonly occur in
sandstones of Upper Triassic (Lower- and Middle-Keuper)
age.
Also within the water absorption, further details are
recognizable: The coefficient of water absorption and the
porosity also show interrelationships to the pore radii dis-
tribution (Fig. 14a) and the different rock types with
varying degrees of alteration (Fig. 14b). Ellipsoids confirm
that sandstones with unimodal equal pore radii distribution
exhibit higher w values in dependence to porosity, whereas
sandstones with unimodal unequal pore radii distribution
show much lower w values. Medium ranges show sand-
stones with bimodal pore radii distribution. This probably
can be attributed to the high content of capillary pores,
which promote the absorption of water in the pore space.
Furthermore, the coefficient of water absorption is highest
for weakly cemented and less compacted quartz arenites. In
contrast, quartz arenites that are more compacted and
cemented show less absorption of capillary water. Sublit-
harenites, which are strongly compacted and cemented,
show low to medium water absorption and low porosity.
The 80 % confidence region of strongly altered sublit-
harenites covers a wide range and overlap with the confi-
dence region of strongly compacted quartz arenites and
sublitharenites. Well-sorted sublitharenites show the high-
est porosity and also a high degree of water absorption.
The ‘‘Houseknecht diagram’’ is used to evaluate which
diagenetic processes reduced the intergranular volume in
order to determine the reservoir quality of a sandstone
(Fig. 15; Houseknecht 1987). In the present paper, the
Houseknecht diagram is used to illustrate the interrela-
tionship between amount and type of cement as well as
intergranular pore volume (minus cement porosity) and
compressive strength. Generally, the compressive strength
increases with decreasing amount of inter granular volume
in dependence on the type of mineral cement (Fig. 15).
Sublitharenites containing a clay cement or pseudomatrix
show a decrease of compressive strength with increasing
intergranular pore volume and a decreasing amount of
cement. Sublitharenites, which are cemented by silica,
carbonate and Fe phases, exhibit the highest strength
(105–150 MPa), an intergranular volume between 20 and
28 % and content of 10–20 % cement. These sandstones
are Carboniferous, Permian and Lower Triassic in age.
Quartz arenites with an intergranular volume of about
20–25 % and a low amount of cement (*5–8 %) show a
compressive strength ranging between 70 and 100 MPa.
Weathering properties
Two of the most important weathering mechanisms in
sandstones are hygric dilatation and salt splitting. In the
literature, the data regarding the weathering properties of
sandstones are commonly based on hygric dilatation under
water saturation and on salt splitting tests with different
types of salt (e.g. sodium chloride, sodium sulphate,
magnesium sulphate). The data used for the diagram in
Fig. 17 are predominantly based on salt splitting tests using
Na2(SO4).
The mean hygric dilatation of all sandstones analysed is
0.49 mm/m (highest values range from 1.2 to 5.21 mm/m;
Fig. 16). Quartz arenites show a very low hygric dilatation
(mean = 0.05 mm/m). Sublitharenites and litharenites
show a dilatation with a mean of 0.59 mm/m. Sorting the
hygric dilatation according to stratigraphic age, sandstones
of the Upper Keuper (Ra¨t) and Cretaceous (Wealden) show
very low dilatation, although these quartz arenites can
contain low amounts of swelling clay minerals (e.g. the
Bad Bentheim and Ro¨hnberg sandstones, Tables 1, 3). This
may be due to the position of clay minerals within the pore
space and also the type of pore radii distribution, which is
almost unimodal equal and does not promote hygric dila-
tation. Distinctly higher values belong to the sandstones of
the Lower Triassic Buntsandstein, which exhibit a mean
hygric dilatation of 0.56 mm/m and a wider scattering of
the first and third quartile. Lower and Middle Keuper
sandstones show a narrow range of hygric dilatation, but
also some very high values of up to 5.21 mm/m, which
reflect litharenites of the Upper Triassic (Lette-
nkohlenkeuper and Schilfsandstein). This extraordinary
high value belongs to the Middle Keuper Hindfelden
sandstone (Table 3), which occurs in the Thuringia Basin.
Sandstones of Cretaceous show with 0.12 mm/m the low-
est hygric dilatation in average.
In salt splitting tests (Fig. 17), 28.3 % of all sandstones
(Group A) analysed show no loss of material up to 100
cycles. 32.6 % of the sandstones (Group B) lose 50 % of
their material in the range of 30–40 cycles. 39.1 % of the
sandstones (Group C) show 50 % material loss already
during the first 15 cycles. Sandstones with the best salt
resistance (Group A; Upper Keuper and Cretaceous,
Table 3) are predominantly composed of quartz arenites.
They seldom fall into Group B or C, such as the Ro¨hnberg
or Cotta sandstone, which is very weakly compacted and
cemented and exhibits high amount of clay cutanes on
detrital quartz grains, respectively, a clayey matrix like
sandtone Cotta. Litharenites are primarily positioned in
Environ Earth Sci (2013) 69:1299–1332 1317
123
1318 Environ Earth Sci (2013) 69:1299–1332
123
Group C, and subordinate in Group B. In contrast, sublit-
harenites show no clear preference to group since every
group is represented. The strongly compacted Weser
sandstones mostly belong to Group A or B. In contrast, the
sandstones of the Lower and Middle Keuper (Schilfsand-
stein, Lettenkohlenkeuper, e.g. SA, SL, HS, IS; Table 3)
mainly fall into Group C and subordinate in Group B.
The interrelationship between the coefficient of water
absorption and porosity in the three groups of salt resis-
tance is documented in Fig. 18a. Sandstones with a high
porosity and a high coefficient of capillary water absorp-
tion show the lowest resistance against salt splitting. In
contrast, sandstones with low porosity and low capillary
water uptake show a very high resistance against salt
splitting. Similar interrelationships are obvious when con-
sidering the compressive strength versus porosity
(Fig. 18b). Sandstones with high compressive strength and
low porosity are most stable against salt splitting, in con-
trast to sandstones with low compressive strength and high
porosity, which show the lowest salt resistance. In Fig. 18c,
four classes of hygric dilatation are defined that show a
distinct overlapping of the confidence ellipsoids. Sand-
stones with low compressive strength and high porosity
primarily belong to the classes of higher hygric dilatation.
Four groups of hygric dilatation can be identified when
the sorption and the amount of micropores are regarded
(Fig. 18d). Samples with different amounts of micropores
and sorption plot in markedly different ellipsoids of 80 %
confidence. Sandstones which show a hygric dilatation of
less than 0.1 mm/m have the lowest amount of micropores.
Nevertheless, they can show a sorption ranging between
0.2 and 1.5 wt%. Sandstones with a hygric dilatation of
0.1–0.5 mm/m have a higher amount of micropores and
water absorption up to 4.5 wt%. Sandstones with a hygric
dilatation of 0.4–5.0 mm/m show the best normal linear
correlation between sorption and the amount of
micropores.
Multicorrelation of sandstone properties using principal
component and cluster analysis
The principal component analyses consider the porosity,
capillary water absorption, compressive strength, salt
resistance as well as hygric dilatation data (data from
present study, Bartelsen 2008; David 2006; Seidel 2004,
Fig. 19). An orientation of variables within the diagram
confirms a positive linear correlation between compressive
strength and the number of cycles in salt splitting tests as
well as between capillary water uptake and porosity.
Between compressive strength, salt resistance and capillary
water uptake as well as porosity, a negative correlation is
well expressed. Hygric dilatation and porosity show no
correlation to each other, whereas a positive correlation of
compressive strength to salt resistance is obvious (Table 4).
The three groups of different pore radii distribution
(Fig. 19a) form clear clusters in the principal component
analysis. The variance of sandstones with unimodal equal
pore radii distribution is characterized by a high capillary
water uptake and porosity. In contrast, the variance of
samples with unimodal unequal pore radii distribution is
concentrated in the field of salt resistance and compressive
strength. The sandstone data exhibiting a bimodal pore
radii distribution show more scattering and are character-
ized by a high hygric dilatation and lower compressive
strength.
Adding lithotypes to the principal component analysis
(Fig. 19b), it is obvious that quartz arenites show a uni-
modal equal pore radii distribution. Strongly compacted
and cemented sublitharenites–litharenites are characterized
by variable data with regard to salt weathering, compres-
sive strength and hygric dilatation. Most of these sand-
stones show high compressive strength and are resistant
against salt splitting, but the hygric dilatation can vary
significantly. These sandstones predominantly show a
unimodal unequal pore radii distribution. In contrast, sub-
litharenites–litharenites with a high content of unstable
lithic fragments and/or pseudomatrix show a higher hygric
dilatation and high porosity with a well-developed ten-
dency for capillary water absorption. These sandstones
primarily show a bimodal pore radii distribution.
By including micropores, the primary intergranular pore
volume (minus cement porosity), unstable lithic fragments
and the sorption into the principal component analysis
(Fig. 19c) further interrelationships can be recognized. The
hygric dilatation shows a positive correlation to the amount
of unstable lithic fragments and sorption. Subordinately, an
interrelationship between the amount of micropores and
degree hygric dilatation can also be observed. The present
porosity always shows a positive linear correlation to the
primary intergranular pore volume (minus cement poros-
ity) in the sandstones analysed.
Fig. 12 Regression diagrams between different petrophysical prop-
erties. The blue ellipsoid predicts an 80 % probability of the data. The
black dashed line next to the regression line represents the ratio
between two properties in some cases. a Porosity versus bulk density
[n = 305; Grimm (1990), Kirsten (2009), David (2006), Holzwarth
(1996), Mosch and Siegesmund (2007), Wenzel and Ha¨fner (2003),
and this study], b capillary water uptake versus compressive strength
[n = 305; David (2006), Holzwarth (1996), Kirsten (2009), Mosch
and Siegesmund (2007), and this study], c water uptake under
atmospheric pressure and under vacuum versus porosity [n = 214;
Kirsten (2009), David (2006), Grimm (1990), and this study],
d ultrasonic velocity versus porosity [n = 89; Fitzner (1970), Kirsten
(2009), and this study], e compressive strength versus ultrasonic
velocity [n = 89; Fitzner (1970), Kirsten (2009), and this study],
f compressive and flexural strength versus porosity, g flexural strength
versus compressive strength and h compressive strength versus bulk
density
b
Environ Earth Sci (2013) 69:1299–1332 1319
123
Weakly compacted quartz arenites with minor amounts
of cement are characterized by high coefficients of water
absorption, a high amount of primary intergranular pores
and high porosity as well as a small hygric dilatation, low
sorption and minor amounts of micropores. These sand-
stones commonly show moderate to low salt resistance and
compressive strength. The extraordinary Bad Bentheim
sandstone exhibits a high capillary water absorption and
high porosity and is characterized by medium compressive
strength and high salt resistance.
Strongly compacted and cemented quartz arenites are
characterized by high compressive strength and salt resis-
tance and low amounts of intergranular pore volume as
well as a low coefficient of water absorption, small
amounts of micropores and a very low sorption. Strongly
compacted and cemented sublitharenites–litharenites are
characterized by small amounts of intergranular pores,
minor water absorption and a low degree of hygric dila-
tation combined with a low amount of micropores.
Strongly altered sublitharenites–litharenites show a high
degree of water absorption and many micropores as well as
a low amount of intergranular pores. The Hindfelden
sandstone exhibits an extraordinarily high degree of hygric
dilatation, many micropores and unstable lithic fragments,
and low salt resistance and compressive strength.
In the cluster analysis shown in Fig. 19d, the different
lithotypes based on similar material and weathering prop-
erties show a very good grouping (I, II, III, IV, and V) and
indicate a well-expressed weathering resistance. Sand-
stones showing low salt resistance and/or distinct hygric
swelling are found in the pink ellipsoid (I). They consist of
strongly altered sublitharenites–litharenites (e.g. the
Fig. 13 Diagrams showing the interrelationship between strength and
porosity to a pore radii distribution (n = 76), b type of grain contacts
(n = 120), c degree of alteration and cementation (n = 120) and
d amount of unstable lithic fragments (n = 120). Data origin from
this study, Bartelsen (2008), David (2006), Fitzner (1970), Kirsten
(2009) and Seidel (2004)
1320 Environ Earth Sci (2013) 69:1299–1332
123
Ingersleben and Hindfelden/Middle and Lower Keuper
sandstones), but also weakly compacted quartz arenites (of
Cretaceous and Lower and Triassic/Buntsandstein, e.g.
sandstone Ru¨then, Hilssandstein, Uder and Velpke).
Sandstones that are found in the orange ellipsoid are mostly
quartz arenites and sublitharenites–litharenites, which are
weakly compacted and cemented, medium to low strength,
show a porosity [20 % and a high coefficient of water
absorption. Furthermore, they are characterized by a
medium to low weathering resistance, since they only show
weak salt resistance. Sandstones belonging to this group
are e.g. Ro¨hnberg and Cotta. In Group III, altered sublit-
harenites–litharenites and weakly compacted quartz are-
nites can be found. They are characterized by porosity
between 10 and 20 %, medium strength as well as medium
weathering resistance, which is confirmed by their affilia-
tion in salt Group B and low hygric dilatation. Sandstones
belonging to this group are Schleerieth, Sander (Upper
Triassic/Lower and Middle Keuper) and Hockeln (Upper
Triassic, Rha¨t). Sandstones of Group IV include the sub-
litharenites–litharenites, which are strongly compacted and
cemented such as the Weser sandstones and also the
Permian Tambach and Bebertal sandstones. They are
characterized by a well-pronounced salt resistance and also
a distinct hygric swelling. Within Group V, strongly
compacted quartz arenites and sublitharenites are found,
which are characterized by low porosity, high strength, low
water absorption and, therefore, a very well-developed
weathering resistance, since they only show little hygric
dilatation and high salt resistance. Sandstones belonging to
Fig. 14 Coefficient of water absorption dependent on porosity.
Ellipsoids mark confidence regions of 80 % probability of a pore
radii distribution (n = 92, Kirsten 2009; David 2006 and data of this
study) and b rock types and degree of alteration, sorting and
compaction [n = 87, Fitzner 1970; Kirsten 2009 and data of this
study]
Fig. 15 Houseknecht diagram
with selected sandstones and
data of compressive strength
Environ Earth Sci (2013) 69:1299–1332 1321
123
this group are, e.g. Seebergen, Arholzen, Obernkirchen and
Ibbenbu¨ren. Only sample Bad Bentheim does not fit in any
group and is separated from the remaining sandstones.
To summarize, the sandstones which belong to Group IV
and V can be defined as weathering resistant, whereas sand-
stones of Group I and II only show weak weathering resistance.
Group III includes samples that can be defined as medium
resistant.
Discussion and conclusions
The statistical analyses from the combined petrographical,
petrophysical and fabric data of the sandstones with dif-
fering geological age and composition show that multiple
interlocking parameters control their material properties. In
the sandstones analysed, the univariate data distribution
highlights the interrelationships between compressive
strength and amount of porosity with lithology and strati-
graphic position. Furthermore, a general decrease of
porosity in dependency to burial time and depth can be
observed, which confirms the study of Koch and Sobott
(2008) and the modelling of Scherer (1987) and Waples
(2002). The sorting as textural parameter of the sandstones
simultaneously affects the porosity. In contrast, no linear
correlation is reflected by the average grain size to poros-
ity, which was also found out by studies of unconsolidated
sands (Beard and Weyl 1973).
The compressive strength shows interrelationships to the
pore radii distribution, the type of grain contacts, the pri-
mary intergranular volume (minus cement porosity) as well
as to the degree of alteration, amount of cementation and
content of unstable lithic fragments. Furthermore, a well-
expressed interrelationship between the amount of
mechanical stable grains and the compressive strength can
be confirmed. Moreover, a high compressive strength is
associated with a longer time and increasing depth of
burial, which also correlates with a lower amount of
porosity. After Fu¨chtbauer (1967) and Houseknecht (1984),
a negative relationship between grain size and volume of
quartz cement exists. Regarding a stabilizing effect by
quartz cementation on sandstones, also a positive correla-
tion between grain size and strength should be observable.
However, a relationship between average grain size and
strength was not detected.
The amount, size, and geometry of the pores signifi-
cantly affect the weathering resistance of a natural building
stone, since they trigger the water balance (Fitzner and
Snethlage 1982; Putnis et al. 1995; Putnis and Mauthe
2000; Ruedrich et al. 2011). As documented in this study,
the properties such as capillary water absorption, water
absorption under normal conditions and under vacuum, as
well as the resulting saturation degree are closely related to
the porosity characteristics of a given sandstone. The
sorption depends significantly on the amount of micropores
(Peschel 1983; Klopfer 1985; Siegesmund and Du¨rrast
2011). Moreover, the pore radii distribution defines the
kind of water transport (Ruedrich and Siegesmund 2006;
Fig. 16 Hygric dilatation of all sandstones analysed and from
literature of Kirsten (2009), David (2006), Holzwarth (1996) (Upper
Keuper, Lower and Middle Buntsandstein, Lower and Middle
Keuper; quartz arenites, lithic arenites and sublitharenites)
Fig. 17 Sandstones of different sandstones [n = 48; David (2006),
this study] with salt resistances classified into Group A (no material
up to 100 cycles), Group B (50 % material loss up 30–40 cycles) and
Group C (50 % material loss until the 15th cycle)
1322 Environ Earth Sci (2013) 69:1299–1332
123
Siegesmund and Du¨rrast 2011) and, therefore, significantly
influences weathering resistance of sandstones (Fitzner and
Snethlage 1982).
Compressive strength, porosity and pore radii distribution
can be defined as general key parameters for the prediction of
salt resistance and hygric dilatation. However, these param-
eters are the result of the primary mineral composition
(compositional maturity) and the diagenetic development
(lithification; type and intensity of cementation; formation of
secondary porosity) Stu¨ck et al. (2011). Based on these
parameters, a model for the interpretation of the diagenetic
development of the different sandstone types and their mate-
rial properties will be discussed below. The goal of this model
is to characterize the development of pore space geometry and
to predict/estimate their possible weathering behaviour.
Diagenetic sequences in different sandstone types
and the development of pore space geometry
and compressive strength
Transport mechanisms and the physicochemical eogenetic
conditions in the depositional milieu (water chemistry;
intrastratal solution) of detrital grains are the basic
parameters for the following diagenetic pathways during
eogenetic alterations and finally trigger the material prop-
erties markedly. The initial porosity is fixed with deposi-
tion of the unconsolidated sediment, as expressed by the
textural and compositional maturity. Houseknecht (1988)
postulates an initial porosity for subarkosic wackes
deposited in meandering river systems of 60 %, and for
subarkoses from braided rivers with 40 %. Analyses of
Fig. 18 Diagrams showing the interrelationships between selected
petrophysical parameters and salt resistance: a coefficient of water
absorption, porosity and groups of salt resistance (n = 48; David
2006 and own), b compressive strength, porosity and salt resistance
(n = 48; David 2006 and own samples), c compressive strength,
porosity and hygric dilatation (David 2006; Kirsten 2009; Holzwarth
1996 and own samples) and d sorption (weight increase at 100 % rel.
air humidity), content of micropores and hygric dilatation. Ellipsoids
illustrate the 80 % confidence regions
Environ Earth Sci (2013) 69:1299–1332 1323
123
artificial grain mixtures of natural sands show that poorly
sorted sands have a porosity of 27.9 %, weakly sorted
sands 30.7 %, moderately sorted 34 %, well-sorted 39 %
and for very well-sorted sands 40.8 % (Beard and Weyl
1973). These data are consistent with the data of sandstones
analysed in the present study. Well-sorted quartz arenites
show the highest intergranular pore volume of 25–30 %,
whereas moderately sorted sandstones show a lower
intergranular pore volume (15–20 %). However, where
unstable lithic fragments are altered to pseudomatrix in
sandstones, the intergranular volume and secondary
porosity can increase greatly. After Schmidt and McDonald
(2012), secondary porosity is also caused by mesogenetic
leaching of carbonate minerals such as calcite, dolomite
and siderite. Besides the primary composition, clay infil-
tration or Fe-oxide rims formed during eodiagenesis have a
marked influence on the later grain overgrowth by silica
(Cecil and Heald 1971) and the formation of special clay
mineral cements. When clay rims observed are thick
(2–5 lm), quartz overgrowth is hindered. If Fe–organic
complexes are present during the mesodiagenetic burial
stage, authigenic chlorite cement rims can be formed
around grains. If Fe–organic complexes or other Fe phases
are missing, illite cement rims are commonly formed
(Surdam et al. 1989). Besides the influence of clay minerals
on the later cementation, the content of clay minerals in
general affects the acoustic velocities and also the com-
pressive strength of sandstones (Kowallis et al. 1984; De
Martine et al. 1976; Han et al. 1986). In general, increasing
Table 4 Components and loadings of principal component analysis
of Fig. 19a, b
Comp. 1 Comp. 2 Comp. 3 Comp. 4 Comp. 5
0.57 0.23 0.14 0.03 0.02
Loadings
Porosity -0.55 0.255 0.796
Hygric
dilatation
-0.17 0.71 -0.64 0.111
Compressive
strength
0.542 0.172 -0.25 -0.69 0.366
Cap. water
absorption
-0.33 0.586 -0.61 -0.11 -0.41
Salt resistance 0.521 0.234 -0.34 0.710 0.228
The loadings are documented for each rock property
Fig. 19 Diagrams showing
principal component analyses in
dependency on a pore radii
distribution, b lithotype
(abbreviations: qtz quartz
arenite, sublith. sublitharenites,
lith. litharenites) and
c additional properties like
sorption, micropores, unstable
lithic fragments and
intergranular volume in
interrelationship to different
lithotypes. d Cluster analyses
with clusters based on
petrophysical properties of
different sandstones. From the
left to the right, the weathering
resistance increases (for
explanation see text; analyses
include own data as well as data
of Bartelsen (2008), David
(2006), Kirsten (2009), Seidel
(2004))
1324 Environ Earth Sci (2013) 69:1299–1332
123
clay contents result in decreasing ultrasonic velocities
independent of the degree of consolidation.
Grains can be differentiated into mechanical and
chemically stable/unstable components (McBride 1963,
Fig. 20). Therefore, during compaction under burial con-
ditions, the content of ductile grains plays an important role
for the further development of the pore space. Large dif-
ferences in the porosity during burial diagenesis are gen-
erated by mechanical and chemical compaction (Surdam
et al. 1989; Hartmann et al. 2000). In comparison to feld-
spathic and lithic sandstones, quartz arenites require deeper
burial conditions for compaction (porosity decrease) to the
same degree (Fig. 21, Dickinson 1985). Regression anal-
ysis between compressive strength and amount of porosity
shows that the content of unstable lithic fragments mark-
edly influences the data documented in the strength–
porosity plot. This confirms that the higher the content of
unstable lithic fragments, the lower is the grain contact
between stable grains probable. Moreover, the type of grain
contacts has an influence on compressive strength. Sand-
stones with predominantly concave–convex and sutured
grain contacts exhibit a higher compressive strength than
sandstones with pointed and elongated grain contacts.
Compressive strength also varies within the groups of grain
contact due to the different kinds and intensity of cemen-
tation. In addition, the kind of unstable lithic fragments is
also important for later cementation. If volcanic detritus is
present, hydration reactions are favoured and the formation
of authigenic feldspar is common. In contrast, abundant
chert fragments can inhibit quartz cementation and thus
preserve porosity (Surdam et al. 1989).
Furthermore, our results also show that porosity and
compressive strength are dependent on the burial time. As
Scherer (1987), Waples (2002) and Koch and Sobott
(2008) documented, sandstones of the same sorting and the
same quartz content which were buried into the same depth
can show different amounts of porosity that is related to the
different time of burial.
These interrelationships are also reflected by all the data
of sandstones recently analysed as schematically summa-
rized in Fig. 22. Pure quartz arenites are commonly well
sorted (high maturity), have a primary porosity of
39–40 %, a unimodal pore radii distribution and reveal
capillary pores of 0.1–10 lm radius. During the closer
packing of grains, pressure solution starts at the point
contacts (chemical compaction) and porosity further
decreases. Depending on a possible infiltration of clay
during eogenesis or not, a quartz dissolution/cementation
reaction takes place. If there is no clay infiltration, the
sandstone may be more intensively compacted and a high
contact thickness (Fu¨chtbauer 1988) occurs, which is
defined by the grain contact length in the entire sandstone
and depends on the type of grain contact. The potential for
quartz pressure solution is given in clay-bearing sandstones
since K2O increases the pH value. This promotes the
solution of quartz (Fu¨chtbauer 1974) and therefore also
influences the porosity modification. The pore radii distri-
bution commonly reveals the same unimodal equal distri-
bution as in the beginning diagenetic stages. Nevertheless,
the average pore size becomes smaller because of the
reduction of larger gusset pores by quartz cementation. The
pore radii distribution of sandstones, which are much more
Fig. 21 Porosity relationship for sandstones of different composition.
Modified after Dickinson (1985)
Fig. 20 Simplified illustration of relationship between lithology
(Q quartz, F feldspar, L lithoclast), mechanical properties during
compaction and pore space development. Arrows show trend of
porosity decrease through compaction and decrease of chemical and
mechanical stability with increasing amount of unstable lithic
fragments and resulting weathering resistance (Modified after
McBride 1963)
Environ Earth Sci (2013) 69:1299–1332 1325
123
1326 Environ Earth Sci (2013) 69:1299–1332
123
strongly compacted, can shift to unimodal unequal and
smaller capillary pores. In the same way, the water
absorption increases due to the presence of more capillary
pores.
These interrelationships are observable in Upper Trias-
sic (Rha¨t; e.g. RO¨, SE, Tables 1, 3) and Cretaceous (e.g.
BB, RU¨, OB, VE, Tables 1, 3) quartz arenites investigated
in this study. Dependent on burial time and sorting, the
porosity is between *15 and 30 %. Sandstones belonging
to the Upper Triassic exhibit a porosity of around 15–20 %
(e.g. SE, Tables 1, 3), whereas younger sandstones
belonging to the Cretaceous show a porosity ranging from
25 to 30 %. According to Blo¨cher et al. (2007), a homo-
geneous pore size distribution results in a higher com-
pressive strength. This can be transferred to the quartz
arenites that show a unimodal equable pore radii distribu-
tion. Because no unstable lithic fragments are present,
these two sandstone types of Upper Triassic and Creta-
ceous show a medium to high compressive strength
(70–150 MPa). Within the glauconite-bearing Cretaceous
sandstones, furthermore a weak negative relationship to
compressive strength is observable (e.g. HI, CO, RU¨).
Further impact on strength can have a clayey matrix, which
distinctly decreases the rock strength like in case of sand-
stone Cotta. Simultaneously, if early clay infiltration took
place and clay rims are formed around detrital grains,
quartz overgrowths are inhibited and micropores occur in
increasing amounts, which both clearly decreases the
compressive strength (e.g. RO¨, Tables 1, 3). Here, the kind
of infiltrated clay minerals are also important: While pore-
filling kaolinite reduces porosity and has little effect on
permeability, lining the pores with illite has little effect on
porosity, but permeability becomes considerably reduced
by blocking pore throats (Stalder 1973; Tucker 2001).
According to Stonecipher et al. (1984), low permeability
promotes cementation. In contrast, in sandstones with high
permeability, solutions do not remain in pore spaces long
enough to build local concentration that promotes precip-
itation of cement. However, in both cases the pore size
distribution through clay minerals is modified from uni-
modal equal to unimodal unequal. This results in a lower
strength (50–80 MPa), since the heterogeneity of pore size
distribution increases and the porosity (*25 %) is pre-
served by prohibited cementation of silica or calcite by the
presence of clay minerals. Because of the effect of clay
mineral cements, which can be compared to clay matrix
and/or pseudomatrix, the grain contacts of detrital grains
are predominantly point contacts.
When the content of unstable lithic fragments and/or
feldspar increases, such as in subarkoses, sublitharenites
and litharenites, the development of pore space and com-
pressive strength is independent of each other (Fig. 22). In
these sandstones, the compositional and textural maturity is
much lower than in quartz arenites. Furthermore, the ratio
between textural to compositional maturities also signifi-
cantly influences the quality of the latter natural building
stone. The so-called Weser sandstones (Lower Triassic;
Tables 1, 3: samples KA, LO, AH) are examples of this
sandstone type. The primary porosity of the moderate- to
well-sorted sandstone is 34–39 %. The pore radii distri-
bution is unimodal unequal. During eogenesis, haematite
impregnation occurs. With advancing burial compaction,
feldspar alteration and quartz dissolution commonly takes
place. Subsequent to quartz cementation, authigenic clay
minerals are formed due to the alteration of feldspars or
unstable lithic fragments and also carbonate cementation
occurs. The intergranular pore space decreases markedly
due to compaction and cementation. The alteration of
feldspar and unstable lithic fragments create a small
amount of micropores. The final pore radii distribution is
unimodal unequal and the compressive strength is depen-
dent on the degree of alteration and cementation that is
medium to high (70–150 MPa). The coefficient of capillary
water absorption is low (*0–5 kg/ms2-), because only a
very small amount of capillary pores is developed and
porosity is mostly low. Sandstones of similar composition
but longer burial history such as the Carboniferous and
Permian sandstones (Tables 1, 3; samples IB, TA, BE)
show a similar pore size distribution, but markedly lower
porosity (up to 10 %) and a higher compressive strength
(70–160 MPa). Due to the lower porosity, also the water
absorption is weakly pronounced.
Sublitharenites and litharenites of the Upper Triassic
Schilfsandstein and Lettenkohlenkeuper show very low
textural and compositional maturity (e.g. HS, IS, SA, SL;
Tables 1, 3). The grains are angular to subangular and
unstable components are moderately altered. The primary
porosity also ranges from 34 to 39 %, because the rock is
moderately well sorted. However, the textural maturity is
characterized by contents of squeezed altered unstable
lithic fragments to pseudomatrix, which distinctly decrea-
ses the textural maturity of these sandstones. After
Fig. 22 Development of pore space and compressive strength of
different sandstone types. After deposition, the primary composition
and sorting rule the (primary) porosity (intergranular volume = IGV)
and the pore radii distribution. Therefore, mature quartz arenites show
higher porosity than more immature sublitharenites–litharenites. This
porosity is modified through clay infiltration, alteration, cementation
and compaction. During eogenesis, clay infiltration can occur that
creates micropores and prohibits quartz overgrowths during later
mesogenesis, which is why the porosity is preserved. Compaction and
cementation decrease the porosity during further diagenesis, and thus
contact thickness can increase, which finally leads to higher strength.
Estimated for each lithotype are grain contact, strength, pore radii
type and porosity. Since the controlling parameters during diagenesis
can vary, the material properties also have to be interpreted as flowing
transitions
b
Environ Earth Sci (2013) 69:1299–1332 1327
123
deposition, porosity decreases through compaction and
silica cementation. Partially, organic fragments parallel to
bedding destabilize the fabric (e.g. HS, IS). The alteration
of (volcanic) clasts to clay minerals and clayey matrix
tends to be squeezed between more rigid grains (pseu-
domatrix), even during the initial stages of diagenesis at
shallow burial. This also prohibits a free circulation of
fluids and leads to restricted cementation. Porosity
decreases through further compaction and the content of
micropores increases greatly because of the altered unsta-
ble lithic fragments. This finally leads to sandstone with a
low compressive strength, because in this rock with het-
erogenous pore radii distribution a stable grain framework
does not exist anymore. The bimodal pore radii distribution
supports high water absorption. The maximum in capillary
pores promotes the capillary water absorption and the
maximum in micropores the sorption. Statistical data of the
different lithologies, fabric properties and type of grain
contact, compressive strength, porosity and pore radii dis-
tribution are documented and summarized schematically in
Fig. 22.
Considering the lateral change of sorting, an increase of
textural maturity and decrease of grain size with increasing
transport distance from the source area should be devel-
oped for each sandstone. Accordingly, a lateral change of
primary composition due to the content of unstable grains,
sorting and primary porosity should be developed. How-
ever, the increase of maturity has to be regarded under very
local conditions with alternating directions and cannot be
transferred by only considering the original source area of
the sediment. For example, the Keuper sandstones of the
Burgsandstein (Vindelizian Keuper) should show an
increase of compositional and textural maturity from SE to
the NW direction. However, these sandstones were pre-
dominantly deposited in channels of alluvial fans, and
therefore they show a highly variable maturity. Thus,
general statements on lateral development of material
properties are not possible in the present study.
Weathering behaviour in dependence to composition
Regression analysis and principal component analyses
allow the prediction of pore space, compressive strength
and water absorption for sandstones of differing lithology.
Weathering behaviour can therefore be estimated because
these parameters show linear correlations to the petro-
physical properties. Although porosity is a key for char-
acterizing the weathering behaviour of the sandstones, the
same amounts of porosity can lead to different resistances
against salt splitting or hygric dilatation. The results of the
salt splitting tests agree with other studies on German
sandstones (Snethlage and Wendler 1997; Fitzner and
Snethlage 1982; Zehnder and Arnold 1989; Ruedrich et al.
2006), where it was documented that sandstones with a
high amount of very small capillary pores and micropores
are highly sensitive to salt splitting. According to Snethlage
and Wendler (1997), the pore space characteristics control
the solution transport and in consequence also the distri-
bution and amount of salt in the rock. Thus, a correlation
between the stability against salt splitting, pore size dis-
tributions and lithotype is well expressed. Furthermore, the
compressive strength is also important for the interpreta-
tion of stability against salt splitting, because the rock has
to withstand increasing crystallization pressures of the salt
(Goudie and Vlies 1997; Scherer 1999), which should not
surmount the degree of internal stability indicated by the
compressive strength. Accordingly, highly mature quartz
arenites, which are weakly compacted (high porosity),
exhibiting a high amount of micropores due to the infil-
tration of clay minerals, have a low-medium compressive
strength (e.g. Ro¨, Tables 1, 3) and show only a moderate
resistance to salt splitting (Group B–C). The Bad Bentheim
sandstone surprisingly is a rock with a very high resistance
to salt splitting and hygric dilatation, although the rock
shows high porosity, medium compressive strength, high
capillary water absorption and swellable clay minerals
(Tables 1, 3). This confirms the special position of sand-
stone Bad Bentheim, illustrated also by cluster analyses
(Fig. 18d). The pore radii distribution with only clear
maxima in large capillary pores probably does not induce
any stress during salt crystallization, since crystal growth
can take place within the free pore space without any
resistance. The homogeneous pore space could be respon-
sible for the medium strength as mentioned above (Blo¨cher
et al. 2007).
Highly compacted and cemented quartz arenites without
the infiltration of clay predominantly exhibit a very high
resistance against salt splitting, because the distribution of
pore radii indicates just a very low amount of micropores
and also only low porosity (e.g. SE, Tables 1, 3). Sand-
stones enriched in unstable feldspars and lithic fragments,
which are highly compacted and cemented, commonly
show a high compressive strength, low porosity and
favourable pore radii distribution (e.g. unimodal unequal),
and therefore are also resistant against salt splitting. Sub-
litharenites and litharenites with common altered rock
fragments and pseudomatrix show the lowest resistance
against salt splitting. The pore radii distribution is char-
acterized by a large amount of micropores caused by the
alteration of unstable lithic fragments (partially squeezed to
pseudomatrix) and a compressive strength which is too low
to withstand the increasing internal pressure by salt crys-
tallization. The lowest salt resistance of the sandstones
analysed are found, e.g. in the Ingersleben and Hindfelden
sandstones of the Upper Triassic (Schilfsandstein, Lette-
nkohlenkeuper) (Tables 1, 3).
1328 Environ Earth Sci (2013) 69:1299–1332
123
Different reasons are discussed for the process of hygric
dilatation. After De La Calle and Suquet (1988), Weiss
et al. (2004), Dixon and Weed (1989) and Moore and
Reynolds (1989), the dilatation is mainly caused by the
presence of expending clay (intracrystalline swelling). Also
discussed is the model of disjoining pressure, comparable
to the intercrystalline swelling caused by interaction within
the electrical double layer at the mineral surfaces (Morales
Demarco et al. 2007; Ruedrich et al. 2011). This process is
limited on very small pore channels. The analyses of hygric
dilatation in the present study confirm that the amount of
micropores triggers the weathering processes markedly,
which confirms a relationship to the presence of clay
minerals, since they create these micropores. Experiments
on clay mixtures prove that swelling clay minerals begin to
expand at a relative air humidity of 45 % (Xiang and
Czurda 1995). However, many sandstones show a volume
increase already at 15 % air humidity (Morales Demarco
et al. 2007; Ruedrich et al. 2011). Because swelling clay
minerals (montmorillonite; mixed layers) only occur in
very minor amounts in the sandstones analysed, other
parameters might also be responsible for the hygric dila-
tation measured. This is underlined in the test with the
quartz arenites (e.g. Cretaceous; Bad Bentheim sandstone).
Both strongly compacted as well as poorly compacted
quartz arenites rarely show hygric dilatation (mean:
0.05 mm/m), although they exhibit swellable clay minerals
(e.g. BB, Tables 1, 3). Probably also the position of the
clay minerals and micropores within the sandstone is
important for hygric dilatation. In contrast, sandstones
exhibiting bimodal pore radii distribution and high alter-
ation of unstable lithic fragments/feldspars or pseudoma-
trix show the greatest hygric dilatation. Maybe the position
of clay in the shape of huge altered lithic fragments pro-
motes water retention behaviour. Eventually also perme-
ability affects the hygric dilatation, since sandstones with
high values of hygric expansion exhibit low permeability
caused by the content of altered unstable lithic fragments.
Sandstones of this type belong to sublitharenites–lithare-
nites of Schilfsandstein and Lettenkohlenkeuper (Upper
Triassic). The highest value of hygric dilatation was mea-
sured in the Hindfelden sandstone (Upper Triassic,
Schilfsandstein). This extraordinary value probably can be
traced back to the combined effects of swellable clay
minerals and the bimodal pore radii distribution with the
high amount of micropores, caused by the presence of
altered lithic fragments. Also, heavily compacted and
cemented Lower Triassic sublitharenites with unimodal
unequal pore radii distribution exhibit distinct hygric
dilatation (e.g. KA, AH, Tables 1, 3). Therefore, we
assume that hygric expansion can be connected with the
interaction of the presence of micropores, the type and
spatial arrangement of clay minerals on surfaces of detrital
grains and on the content of altered unstable lithic frag-
ments (see also discussion in Ruedrich et al. 2011).
Summary—quality catalogue for sandstones
The results of the present study indicate that statistical
analyses of a very large number of sandstones can be used
to predict the material and weathering behaviour of these
natural stones. Multiple parameters affect the material
properties of these heterogeneous rocks. The possible
problems of compiling a data set from many different
sources also have to be kept in mind. Furthermore, petro-
graphical and petrophysical parameters can change locally
within each individual rock and, therefore, the material
properties can also vary considerably.
Linear interrelationships between the basic parameters
of porosity, pore radii distribution and compressive
strength can be established for each sandstone lithotype
defined. Because these parameters correlate with the
weathering properties, a prediction for salt resistance and
hygric dilatation can be made for each sandstone type.
Material properties are dependent on different magnitudes
of influences. Thus, the transition between the different
lithotypes is fluent and, therefore, the weathering behaviour
is common. Figure 23 schematically summarizes the pet-
rographical and petrophysical properties as well as their
resistance against salt splitting and hygric dilatation. In
general, the construction suitability is determined by dif-
ferent classes from the combination of the salt resistance
and hygric dilatation results, in which all other parameters
are incorporated with a very large confidence.
In the present study, two types of quartz arenites and
two types of sublitharenites–litharenites as natural building
stone could be distinguished:
• High mature quartz arenites, which are strongly com-
pacted and cemented with little early infiltrated clay
minerals and mainly sutured concave–convex grain
contact. They exhibit a compressive strength between
70 and 150 MPa, a porosity up to 15 % and a unimodal
unequal pore radii distribution. Due to the low hygric
dilation and high salt resistance (Group A), they can be
defined as weathering resistant.
• Quartz arenites, which are weakly compacted and
cemented, with point-elongated grain contact, partially
containing glauconite grains and high content of early
infiltrated clay minerals. The compressive strength is
between 50 and 80 MPa, the porosity amounts to
*25 % and the pore radii distribution is unimodal
equal. Due to the low hygric dilatation but medium-low
salt resistance, the rock can be defined as moderate–
well weathering resistant.
Environ Earth Sci (2013) 69:1299–1332 1329
123
• Sublitharenites (-litharenites) with a low-medium con-
tent of altered unstable lithic rock fragments and
feldspars, high-medium content of quartz cement as
well as illite and/or kaolinite at grain surfaces. The
compressive strength is 70–150 MPa and medium to
high, the porosity is between 5 and 15 % and the pore
radii distribution unimodal unequal. The salt resistance
belongs to group A–B and hygric swelling amounts to
0.5–1.0 mm/m.
• Sublitharenites (–litharenites) with high content of
altered unstable lithic fragments and/or feldspars, little
quartz cement and weak grain contact between the few
stable detrital grains. The lithic fragments are mostly
squeezed to pseudomatrix. The compressive strength is
30–50 MPa and very low and the porosity is 20–30 %
and very high. Due to the high values of hygric
dilatation and the low salt resistance, this sandstone
type can be defined as not weathering resistant.
Despite the material properties of dimension sandstones
discussed in this paper, the evaluation of a suitable
replacement stone must additionally take into account the
compatibility of a replacement stone to the adjacent rocks.
For example, moisture properties such as capillary and
sorptive water absorption and water saturation of the ori-
ginal and replacement rock should be of the same magni-
tude. A case study considering this interaction of original
and replaced dimension stones is presented by Graue et al.
(2011) for the Cologne cathedral.
Acknowledgments We would like to thank the Deutsche Bun-
desstiftung Umwelt for supporting the long-term PhD fellowship for
H. Stu¨ck (AZ 20008/997). Furthermore, gratitude also goes to Dr.
O. Nenadic and Dr. J. Schuhmacher for the constructively support of
statistics, Dr. R. Dohrmann und Dr. S. Kaufhold for analysing the clay
mineral compositions (BGR), and Christian Gross for editing the
English. Special thanks go to Dr. J. Ruedrich for his help and
unreserved discussions on all experimental results.
Fig. 23 Illustration summarizing a quality catalogue for sandstones.
From left to right, the textural and compositional maturity increases.
Coloured bars indicate the classes of construction suitability. In
general, four different sandstone types with characteristic fabric types
were classified material parameter such as porosity and compressive
strength, as well as the weathering properties characterized by salt
resistance and hygric swelling, are also determined. Salt resistance is
divided into three groups: Group A (no material loss up to 100
cycles), Group B (50 % material loss up 30–40 cycles), and Group C
(50 % material loss up to 15 cycles). The transition between each
lithotype is interpreted as being fluent
1330 Environ Earth Sci (2013) 69:1299–1332
123
Open Access This article is distributed under the terms of the
Creative Commons Attribution License which permits any use, dis-
tribution, and reproduction in any medium, provided the original
author(s) and the source are credited.
References
Aigner T, Bachmann G (1992) Sequence stratigraphic framework of
the German Triassic. Sedimentary Geology 80:115–135
Bartelsen T (2008) Gefu¨geabha¨ngiges Verwitterungsverhalten ton-
mineralfu¨hrender Sandsteine durch hygrothermisch induzierte
Dehnungsprozesse, unpublished diploma thesis, Go¨ttingen
Beard DC, Weyl PK (1973) Influence of texture on porosity and
permeability of unconsolidated sand. AAPG Bull 57:349–369
Beutler G, Schubert J (1987) Fazielle Entwicklung des Mittleren
Lettenkeuper im Thu¨ringer Becken. Z Geol Wiss 15:475–484
Blo¨cher G, Bruhn D, Zimmermann G, McDermott C, Huenges E
(2007) Investigation of the undrained poroelastic response of
sandstones to confining pressure via laboratory experiment,
numerical simulation and analytical calculation. Geological
Society Special Publications, London
Burley SD, Kantorowicz JD, Waugh B (1985) Clastic diagenesis. In:
Brenchley PJ, Williams BJP (eds) Sedimentology: recent
developments and applied aspects. Special publication 18.
Geological Society of London, Bath, pp 89–226
Cecil CB, Heald MT (1971) Experimental investigation of the effects of
grain coatings on quartz overgrowth. J Sediment Petrol 41:582–584
Chitsazian AH (1985) Beziehungen zwischen dem Mineralbestand,
Gefu¨ge und technologischen Eigenschaften der Niedersa¨chsis-
chen ,,Wealden‘‘-Sandsteine (Unterkreide)-Mitteilungen aus
dem Geologischen Institut Hannover, Diss
Choquette PW, Pray LC (1970) Geological nomenclature and
classification of porosity in sedimentary carbonates. Am Assoc
Petrol Geol Bull 54:207–250
David C (2006) Buntsandstein- Bausandstein, Marburger Bausand-
stein unter der Lupe. Zeitschrift der Marburger Geowissens-
chaftlichen Vereinigung e.V., 3, Marburg
De La Calle C, Suquet H (1988) Vermiculite. In: Bailey SW (ed.)
Hydrous phyllosilicates. Reviews in Mineralogy. Mineralogical
Society of America 19:455–496
De Martine DC, Beard D, Danburg JS, Robinson JH (1976) Variation
of seismic velocities in sandstones and limestones with lithology
and pore fluid at simulated in situ conditions: Proceedings of the
EGPC exploration seminar, Cairo
Dickinson WR (1985) Interpreting provenance relations from detrital
modes of sandstones. In: Zuffa GG (ed) Provenance of arenites.
Reidel, Dordrecht, pp 333–361
Dixon JB, Weed SB (1989) Minerals in soil environments. 2nd
edition Soil Science Society of America
Dohrmann CF, Ku¨hn I (2009) Angewandte Statistik fu¨r die biolog-
ischen Wissenschaften. Helmholtz Zentrum fu¨r Umweltfors-
chung-UFZ
Fitzner B (1970) Die Pru¨fung der Frostbesta¨ndigkeit von Naturbau-
steinen. Aachen, Diss
Fitzner B (1988) Untersuchung der Zusammenha¨nge zwischen dem
Hohlraumgefu¨ge von Natursteinen und physikalischen Ver-
witterungsvorga¨ngen. Mitt. Ing.-u.Hydrogeologie, 29, Aachen
Fitzner B, Snethlage R (1982) Einfluß der Porenradienverteilung auf
das Verwitterungs-verhalten ausgewa¨hlter Sandsteine. Bauten-
schutz und Bausanierung 3:97–103
Fu¨chtbauer H (1967) Der Einfluss des Ablagerungsmilieus auf die
Sandsteindiagenese im Mittleren Buntsandstein. Sediment Geol
1:159–179
Fu¨chtbauer H (1974) Zur Diagenese fluviatiler Sandsteine. Geol
Rundsch 63(3):904–925
Fu¨chtbauer H (1988) Sedimente und Sedimentgesteine. 4. Auflage,
1141 S., Stuttgart (Schweizerbart)
Gilbert CM (1954) Sedimentary rocks, p.251–384 in Williams, H.,
Turner, F.J. and Gilbert, C.M. Petrography: San Francisco, W.H.
Freeman & Co., 406 p.
Goudie A, Vlies H (1997) Salt weathering hazards. Wiley, Hoboken
Graue B, Siegesmund S, Middendorf B (2011) Quality assessment of
replacement stones for the Cologne Cathedral: mineralogical and
petrophysical requirements. Environ Earth Sci 63(7–8):1799–1822
Grimm WD (1990) Bildatlas wichtiger Denkmalgesteine der Bun-
desrepublik Deutschland. Bayrisches Landesamt fu¨r Denkmalpf-
lege, 50, Mu¨nchen
Han D, Nur A, Morgan D (1986) Effect of porosity and clay content
of wave velocities in sandstones. Geophysics 51(11):2093–2107
Hartmann DJ., Beaumont EA, Coalson E (2000) Predicting sandstone
reservoir system quality and example of petrophysical evalua-
tion. Predicting reservoir system quality and performance. In:
Hartmann DJ, Beaumont EA (eds) Exploring for oil and gas
traps, Beaumont EA, Foster NH (eds) Treatise of petroleum
geology, handbook of petroleum geology, 1999
Hayes JB (1984) Sandstone porosity evolution. Short course notes,
AAPG, Sandstone diagenesis School
Holzwarth D (1996) Werksandsteine der Region Vorderpfalz (wes-
tlicher Grabenrand des Obberheingrabens) und ihre petrogra-
phischen sowie geotechnischen Eigenschaften. Diss, Mainz
Houseknecht DW (1984) Influence of grain size and temperature in
intergranular pressure solution, quartz cementation, and porosity
in a quartzose sandstone. J sediment petrol 54(2):348–361
Houseknecht DW (1987) Assessing the relative importance of
compaction processes and cementation to reduction of porosity
in sandstones. Am Assoc Petroleum Geol Bull 71(6):642–663
Houseknecht DW (1988) Intergranular pressure solution in four
quartzose sandstones. J Sed Petrology 58:228–246
Houseknecht DW, Pittman ED (eds) (1992) Origin, diagenesis &
Petrophysics of Clay Minerals in Sandstones. Special Publication
47. Society of Sedimentary Geologists, Tulsa
Katzung G (1975) Tektonik, Klima und Sedimentation in der
Mitteleuropa¨ischen Saxon-Senke und in angrenzenden Gebieten.
Z Geol Wiss 3(11):1453–1472
Kirsten H (2009) Herkunft, Eigenschaften und Konservierungs-
mo¨glichkeiten von Lettenkeuper- und Schilfsandsteinen an
Baudenkmalen in Thu¨ringen. Diss, Weimar
Klopfer H (1985) Feuchte. In: Lutz et al.: Lehrbuch der Bauphysik.
Teubner Verlag, Stuttgart, pp 265–434
Koch R, Sobott R (2008) Sandsteine: Entstehung, Eigenschaften,
Verwitterung, Konservierung, Restaurierung. In: Siegesmund,
Snethlage (eds) Schriftenreihe der Deutschen Gesellschaft fu¨r
Geowissenschaften, vol 59, pp 145–174
Koch R, Baier A, Lorenz H, Fritsch A (2003) Sandsteine des Keupers
als Naturwerksteine in und um Nu¨rnberg (Exkursion B am 22.
April 2003). Jber Mitt oberrhein geol Ver NF 85:45–64
Kowallis B, Jones LEA, Wang HF (1984) Velocity–porosity–clay
content: systematics of poorly consolidated sandstones. J Geo-
phys Res 89:10355–10364
Massallam K (1973) Sedimentologie, Petrographie und Diagenese
der Valendis Sandsteine Erdo¨lfeld Leiferde-Gifhorner Trog
(Unterkreide – NW-Deutschland). Diss, Univ. Hamburg
McBride EF (1963) A classification of common sandstones. J Sedi-
ment Petrol 33(3):664–669
Moore DM, Reynolds RC Jr (1989) X-ray diffraction and the
identification and analysis of clay minerals. Oxford University
Press, Oxford
Morad S, Ketzer JM, De Ros LF (2000) Spatial and temporal
distribution of diagenetic alterations in siliclastic rocks:
Environ Earth Sci (2013) 69:1299–1332 1331
123
implications for mass transfer in sedimentary basins. Sedimen-
tology Suppl 1:95–120
Morales Demarco M, Jahns E, Ruedrich J, Oyhantacabal P, Sieges-
mund S (2007) The impact on partial water saturation in rock
strength: an experimental study on sandstones. Zeitschrift der
Deutschen Gesellschaft fu¨r Geowissenschaften 158:869–882
Mosch S, Siegesmund S (2007) Statistisches Verhalten petrophysi-
kalischer und technischer Eigenschaften von Naturwerksteinen.
Z Dtsch Geol Ges 158/4:821–868
Mu¨ller H, Luckert J, Ro¨per D (2002) Sandsteinverwitterung durch
bauscha¨dliche Salze—Ergebnisse der Sa¨ulensanierung des Bel-
vedere auf dem Pfingstberg/Potsdam—Brandenburgische Geo-
wissenschaftliche Beitra¨ge 9:1–15; Kleinmachnow
Paul J (1982) Der Untere Buntsandstein des Germanischen Beckens.
Geol Rundsch 71:795–811
Paul J, Wemmer K, Ahrend H (2008) Provenance of Triassic siliciclastic
sediments in Central Europe. Z dt geol Ges 159:641–650
Peschel A (1983) Natursteine. VEB Deutscher Verlag fu¨r Grundstof-
findustrie, Leipzig
Pettijohn FJ (1975) Sedimentary rocks. Harpers u. Brothers, New York
Pettijohn FJ, Potter PE, Siever R (1987) Sand and sandstone.
Springer, New York
Plein E (1993) Bemerkungen zum Ablauf der pala¨ogeographischen
Entwicklung im Stephan und Rotliegend des Norddeutschen
Beckens. Geol Jb A131:99–116
Putnis A, Mauthe G (2000) The effect of pore size on cementation in
porous rocks. Geofluids 1:37–41
Putnis A, Prieto M, Fernandez-Diaz L (1995) Supersaturation and
crystallisation in porous media. Geol Mag 132:1–13
Rac MV, Chernyshev SN (1970) Klu¨ftigkeit und Eigenschaften
geklu¨fteter Gesteine (russ.), Moskau (Nedra)
Ruedrich J, Seidel M, Kirchner, Siegesmund S (2005) Salzverwitte-
rung, hygrische und thermische Dehnung als auslo¨sende Scha-
densquantita¨ten. In: Siegesmund S, Auras M, Ru¨drich J,
Snethlage R (eds) Geowissenschaften und Denkmalplfege
ZDGG, vol 156/1, pp 59–74
Ruedrich J, Siegesmund S (2006) Fabric dependence of length change
behaviour induced by ice crystallization in the pore space of
natural building stones. In: Fort R, Alvarez de Buergo M,
Gomez-Heras M et al (eds) Heritage, weathering and conserva-
tion. Taylor & Francis Group, London
Ru¨drich J, Bartelsen T, Dohrmann R, Siegesmund S (2010) Building
sandstone integrity affected by the process of hygric expansion.
Environ Earth Sci. doi:10.1007/s12665-010-0767-0
Ruedrich J, Bartelsen T, Dohrmann R, Siegesmund S (2011) Moisture
expansion as a deterioration factor for sandstone used in
buildings. Monument under threat. Environ Earth Sci 63:7–8
Scherer M (1987) Parameters influencing porosity in sandstones: a
model for sandstone porosity prediction. Am Assoc Petroleum
Geologists Bull 71(5):485–491
Scherer GW (1999) Crystallization in pores. Cem Conr Res 29:1347–1358
Schießl P, Alfes C (1990) Festigkeit und Verformung von Sand-
stein—Bedeutung fu¨r die Verwitterungsresistenz und Meßme-
thoden. Denkmalpflege und Naturwissenschaften im Gespra¨ch,
Sonderheft aus der Publikationsreihe des Verbundsforschung-
sprojekts Steinzerfall und Steinkonservierung:10–17, Berlin
Schmidt V, McDonald DA (2012) The role of secondary porosity in
the course of sandstone diagenesis. In: Scholle PA, Schluger PR
(eds) SEPM Society for Sedimentary Geology. Aspects of
Diagenesis, vol 26, pp 175–207
Seidel G (2003) Geologie von Thu¨ringen, Schweizerbart’sche Ver-
lagsbuchhandlung, Stuttgart, 2. Aufl.
Seidel M (2004) Gefu¨geabha¨ngige Verwitterung von Sandsteinen
durch Kristallisation von Eis und Salzen im Porenraum un
published diploma thesis, 159 pp, Go¨ttingen
Siegesmund S, Du¨rrast H (2011) Physical and mechanical properties
of rocks. In: Siegesmund S, Snethlage R (eds) Stone and
architecture. Springer, Berlin, pp 97–225
Snethlage R, Wendler E (1997) Moisture cycles and sandstone
degradation. In: Baer NS, Snethlage R (eds) Saving our
architectural heritage, the conservation of historic stone struc-
tures. Elsevier, Hoboken, pp 7–24
Sobott R, Koch R (2009) Die Qualita¨t von Naturwerksteinen aus dem
Steinbruch Worzeldorf bei Nu¨rnberg—Petrographische, diage-
netische und gesteinsphysikalische Kenndaten (Mittlerer Keuper,
Oberer Burgsandstein). Geol. Bl. NO-Bayern, 59, pp 419–444,
Erlangen
Stu¨ck H, Siegesmund Siegfried, Ruedrich J (2011) Weathering
behaviour and construction suitability of dimension stones from
the Drei Gleichen area (Thuringia, Germany). Environ Earth Sci
63:1763–1786. doi:10.1007/s12665-011-1043-7
Stonecipher SA, Winn RD Jr., Bishop MG (1984) Diagenesis of the
Frontier Formation, Moxa Arch: a function of sandstone
geometry, texture and composition, and fluid flux. In: McDonald
DA, Surdam RC (eds) Clastic Diagenesis: AAPG Memoir 37,
pp 289–316
Stalder PJ (1973) Influence of crystalline habit and aggregate
structure of authigenic clay minerals on sandstone permeability.
Geol Mijnbouw 52:217–219
Strack D, Stapf RG (1980) Ist der Kreuznacher Sandstein des
Rotliegenden a¨olisch oder fluviatil entstanden? Geol Rundsch
69(3):892–921
Surdam RC, Dunn TL, MacGowan DB, Heasler HP (1989) Concep-
tual models for the prediction of porosity evolution with an
example from the Frontier Sandstone, Big-Horn Basin, Wyo-
ming. In: Coalson EB, Kaplan SS, Keighin CW, Oglesby LA,
Robinson JW (eds) Sandstone Reservoirs: Rocky Mountain
Association of Geologists, pp 7–21
Tukey J (1977) Exploratory data analysis. Addison-Wesley. ISBN
18750-201-07616-0
van Brakel J, Modry S, Svata M (1981) Mercury porosimetry: State of
the art. Powder Technol 29:1–12
Van der Plas L, Tobi AC (1965) A chart of judging the reliability of
point counting results. Am J Sci 263:87–90
Waples DW (2002) Evolution of sandstone porosity through time.
The modified Scherer Model: a calculation method applicable to
1-D maturity modeling and perhaps to reservoir prediction. Nat
Resour Res 11:257–272
Wenzel A, Ha¨fner F (2003) Die roten Werksandsteine der Westpfalz.-
IFS Bericht Nr. 15-2003, p 27
Weiss T, Siegesmund S, Kirchner D, Sippel J (2004) Insolation
weathering and hygric dilatation as a control in building stone
degradation. Environ Geol 46(3-4):402-413
Wilson MD (1994) Non-compositional controls on diagenetic
processes. In: Wilson MD (ed) Reservoir Quality Assessment
and Prediction in Clastic Rocks: SEPM Short Course 30,
pp 183–208
Worden R, Morad S (2000) Quartz cementation in sandstones. Special
publication 29, International Association of Sedimentologists.
Blackwell Science, Oxford
Wurster P (1964) Geologie des Schilfsandsteins. Mitt Geol Staats-
instituts Hamburg 33:1–140
Xiang W, Czurda K (1995) Einfluss des Kationenaustausches auf die
Hydratation und Dehydratation von Tonmineralen. Festschrift
zum 60. Geburtstag von Helfried Mo¨ller. Geol. Pala¨ontologische
Mitteilungen Innsbruck, Bd. 20:107–119
Tucker M (2001) Sedimentary petrology—revised. Blackwell Science
Inc, Oxford
Zehnder K, Arnold A (1989) Crystal growth in salt efflorescence.
J Crystal Growth 97:513–521
1332 Environ Earth Sci (2013) 69:1299–1332
123