399
Comparison of permeability testing methods
Comparaison des différentes méthodes sur les tests de perméabilité
Nagy L., akács A. T, Huszák T., Mahler A., Varga G.
Budapest University of Technology and Economics
ABSTRACT: Coefficient of permeability is known as the most variable soil property Its value can vary over an order of magnitude
even in case of relatively homogenous layers. So determining this value is a complicated, complex engineering task. There is a vast
number of laboratory and in situ tests to determine the permeability coefficient. Each method has its own advantages, drawbacks and
limitations, so different methods should be preferred in different situations. The permeability coefficients of a sandy silt and a silty
sand layer have been determined by means Khafagi probe, Menard probe, water filtration method, constant head laboratory test and
falling head laboratory test. The permeability coefficients have also been estimated by the equation proposed by Hazen (1895). The
measured values are summarized and evaluated in the paper. Special emphasis is given on the reliability of the methods, on the
capability to sense the layer boundaries and their estimation accuracy.
RÉSUMÉ : Le coefficient de perméabilité est connu comme étant la propriété du sol la plus variable. Sa valeur peut varier, même
dans le cas de couches relativement homogènes. Ainsi, la détermination de cette valeur est une question difficile, une tâche
d’ingénierie complexe. Il existe de nombreux laboratoires et des tests
in situ
pour déterminer le coefficient de perméabilité. Chaque
méthode a ses avantages, ses inconvénients et ses limites, ainsi certaines méthodes doivent être privilégiées en fonction du contexte.
Les coefficients de perméabilité d’un limon sableux et d’une couche de sable silteux ont été déterminés par la méthode Khafagi, par
l’essai pressiométrique Ménard, par la méthode constante et tomber test de perméabilité à la tête. Les coefficients de perméabilité ont
également été estimés par l’équation proposée par Hazen (1895). Les valeurs mesurées sont résumées et évalués dans le document.
L’accent est mis sur la fiabilité des méthodes, sur la faculté de détecter les limites des couches et sur la précision de leur estimation.
KEYWORDS: coefficient of permeability, laboratory test, in situ test
1 INTRODUCTION
Coefficient of permeability (also known as hydraulic
condictivity, denoted by ‘k’) is a highly variable soil property.
Previous studies have shown that its coefficient of variation can
be as high as 240 % (Lumb, 1966., Uzielli, 2008., Mlynarek,
2010.). Additionaly the chosen tetsting method has also high
influence on measured results.
The two main factors that determine the order of magnitude
of the permeability coefficient are: grain size and cleavage
(secondary interstices). These two properties can already have
significant spatial variability, but other influencing factors make
the determination of permeability coefficient even more
complex. The impact of the factors listed below is inferior, but
still not insignificant:
• grain shape and orientation,
• quantity and connection of interstices,
• uniformity coefficient,
• water content and saturation conditions before seepage
begins,
• the properties of the passing liquid (water),
• hydraulic conditions (hydraulic gradient, Reynolds number
etc.),
• transient phenomena (migration, wash-out and wash-in of
grains).
Section S3 of Annex S to EUROCODE 7: Geotechnical
Design Standard highlights the role of saturation, which may
cause a change of up to three orders of magnitude in the
coefficient of permeability of certain soil types.
It is fair to say therefore that the coefficient of permeability
of soils can depend on a large number of factors of different
character, which is why general relationships (formulas or
graphs) based on a few simple quantities are not expected to
provide accurate k values. Based on these considerations, it is
not a good practice to use values taken from tables of universal
validity. No one can guarantee, for instance, that soils with the
correlation feature I
p
=30% have identical coefficients of
permeability at all sites.
1.1
Background
There is no method specified either as a Hungarian Standard
or in a Technical Guideline for calculating the coefficient of
permeability. Coefficient of permeability values can be
determined by on site or laboratory measurements or indirectly
from empirical correlations based on grain size distribution.
Small as it is, even a country such as Hungary has failed to
come to a common understanding about the test.
Kézdi (1976) expresses a preference for laboratory tests for
determining the coefficient of permeability and indeed the
following laboratory methods are available for determining the
value of k:
• by constant water head test
• by falling water head test,
• by capillary permeability test, and
• from a consolidation test.
Rózsa (1977) rejects the laboratory method and recommends
pumping from a well to determine the coefficient of
permeability:
"The coefficient of permeability is one of the physical
properties that cannot be determined at the required accuracy
using laboratory methods. Frequently, a 10-50 fold accuracy of
Tabácks A.,
