Document Actions

ijege-15_02-francani-et-alii.pdf

background image
5
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
DOI: 10.4408/IJEGE.2015-02.O-01
V
incenzo
Francani
(*)
, P
aola
Gattinoni
(*)
, M
arc
Schöttler
(**)
, S
ara
Mottini
(***)
,
D
ario
S
cott
riGaMonti
(***)
& M
ariangela
Vitiello
(***)
(*)
Politecnico di Milano - Dipartimento Ingegneria Civile e Ambientale (D.I.C.A.) - 20133 Milan, Italy
(**)
PHREALOG - Rheinallee 88 - 55120 Mainz, Germany
(***)
Tethys s.r.l. - Viale Lombardia 11 - 20131 Milan, Italy
ApplicAtion of borehole flow meAsurements to chArActerize
hydrAulic heterogeneities And their impAct
on the locAl groundwAter flow network
extended AbstrAct
il presente studio sperimenta un metodo d’indagine per l’identificazione e la caratterizzazione delle eterogeneità idrauliche presenti
in sistemi acquiferi complessi. l’interesse scientifico ed applicativo per la tematica deriva dal fatto che le modalità di implementazione
di queste eterogeneità all’interno dei modelli matematici, usualmente utilizzati per simulare il comportamento e l’evoluzione dei sistemi
acquiferi, condizionano fortemente i risultati della modellazione stessa. in effetti, il miglioramento del dettaglio e della precisione con il
quale si tiene conto delle eterogeneità all’interno del dominio dà risultati molto positivi, soprattutto nel caso di spiccata eterogeneità lungo
il profilo verticale e per analisi a scala di grande dettaglio (sito specifiche).
obiettivo del presente studio è quindi quello di verificare l’opportunità di integrare i metodi d’indagine tradizionale (log stratigrafici,
prove di portata e monitoraggio piezometrico) con misure di flusso in foro tramite flowmeter, in grado di completare il quadro conosci-
tivo locale attraverso l’individuazione e la caratterizzazione delle eterogeneità. Più in dettaglio, nel presente studio si sono eseguite mi-
sure di flusso sia orizzontale che verticale in foro; tali misure sono state condotte su un caso di studio nel quale l’eterogeneità geologica
dell’acquifero (acquifero multistrato caratterizzato da una trasmissività dell’ordine di 0.04 m
2
/s) si somma alla presenza di strutture antropi-
che interrate, che rendono il sistema particolarmente complesso da caratterizzare con le tecniche di indagine tradizionali.
in particolare le misure di flusso verticale sono state condotte sia in condizioni di flusso indisturbate sia in presenza di pompag-
gio e hanno permesso una dettagliata ricostruzione della successione idro-stratigrafica, con l’individuazione di vari livelli ad elevata
permeabilità nella porzione inferiore di una unità precedentemente ritenuta omogenea, e soprattutto di un livello acquifero confinato
in grado di originare flussi diretti verso l’alto. nei livelli a maggiore permeabilità sono poi state eseguite delle letture piezometriche
tramite cluster mirati alla definizione del carico idraulico alle profondità nelle quali i test con flowmeter verticale avevano individuato
le maggiori portate defluenti.
infine, le misure di flusso orizzontali sono state condotte nei livelli a maggiore permeabilità precedentemente individuati sia nei pi-
ezometri di monitoraggio classici (caratterizzati da lunghe fenestrature) sia nei cluster (aventi fenestrature molto più ridotte); queste ultime
misure si sono dimostrate molto più significative al fine di caratterizzare le eterogeneità dell’acquifero ed hanno fornito valori di velocità
secondo Darcy dell’ordine di 10
-4
m/s, permettendo così di quantificare il flusso attraverso i livelli maggiormente produttivi dell’acquifero,
anche in conseguenza dell’interazione con le strutture interrate.
i dati e le informazioni ottenuti dai test con flowmeter sono stati utilizzati per aggiornare il modello concettuale del sistema acquifero,
inizialmente elaborato sulla base della campagna d’indagine tradizionale. i due modelli concettuali (quello iniziale e quello aggiornato)
sono poi stati confrontati tramite l’implementazione di due differenti modelli numerici di flusso 3D, con l’impiego del codice di calcolo
alle differenze finite MoDFloW.
i risultati delle simulazioni hanno evidenziato come alla scala locale i metodi di indagine tradizionale non siano in grado di caratteriz-
zare l’acquifero con un sufficiente grado di dettaglio, in quanto la presenza di spiccate eterogeneità lungo il profilo verticale porta alla
formazione di livelli acquiferi in pressione lungo gli strati a maggiore permeabilità, che sovrapponendosi agli effetti indotti dalla presenza
delle strutture interrate influenzano fortemente la dinamica del sistema acquifero, generando localmente significativi flussi verticali aventi
gradiente dal basso verso l’alto.
lo studio effettuato consente di giungere ad alcune importanti conclusioni; in particolare risulta confermato che l’esecuzione di misure
di velocità in foro in punti opportunamente scelti può migliorare significativamente il modello concettuale del sistema acquifero, consen-
tendo una ricostruzione del campo di flusso a scala locale di maggiore dettaglio ed affidabilità non solo in condizioni naturali, ma anche
e soprattutto in presenza di strutture antropiche interrate, che esercitino la funzione di barriere impermeabili capaci di deviare in modo
significativo il flusso idrico.
background image
V. frAncAni, p. gAttinoni, m. schöttler, s. mottini, d.s. rigAmonti & m. Vitiello
6
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
AbstrAct
the focus of this study is to examine the use of borehole flow
measuring procedures, in order to support conventional methods
and information used to characterize heterogeneous aquifers. in
general stratigraphic logs, data from pumping tests and piezo-
metric monitoring are used. in this study a combined approach
of vertical and horizontal borehole flow measurements and pi-
ezometric monitoring is described, to complete and merge avail-
able data for the accurate modeling of a complex aquifer. Ver-
tical thermo-flowmeter logs allowed the detection of noticeable
heterogeneities: a leaky aquifer, an upward directed leakage and
several layers with higher hydraulic conductivity in the lower part
of a hydrostratigraphic unit that appeared almost homogeneous.
horizontal in-hole groundwater flow measurements allowed for
quantifying the flow in the more productive layers at the bottom
of the aquifer, with observed Darcy velocity up to 10
-4
m/s
in some
measuring points. a numerical model was applied to analyze the
major differences between a conceptual model of the aquifer ob-
tained by the use of conventional methods of characterization and
a model based on the here proposed approach.
K
eywords
: groundwater velocity, conceptual models, heterogeneity,
flowmeter, numerical modeling, in-hole flow measurements
introduction
Volcanic and alluvial sediments consist of layers of various
thickness, partly with large areal extension, characterized by dif-
ferent granulometry and usually a slightly sloping inclination (the
slope often ranges between 0.1% and 5%). the presence of lenses
with different dimensions and thickness, the occurrence of buried
riverbeds and strata acting as aquitard design a groundwater flow
network with preferential groundwater flow paths and directions.
in general, structural variations of sediments on a regional scale
are due to the alternation of deposition and erosion phases during
the formation of the aquifer. each sedimentation phase is associ-
ated to a particular granulometry and different layers are separated
by irregular erosion surfaces. in Fig. 1 the structure of the alluvial
aquifer of the Milan area is shown as an example. on a local scale,
different shapes and hydraulic conductivities of sediment lenses
and layers take significant influence on the flow network.
the evaluation of the continuity and hydraulic conductiv-
ity of layers and lenses in alluvial and volcanic aquifers is of-
ten difficult owing to the lack of suitable data. nevertheless, the
characterization of the mutual role and productivity of different
strata and their significance in the hydraulic network can play an
important role in understanding the actual flow conditions on a
local scale.
it is known that a vertical flow can be naturally induced in a
borehole that cuts across hydraulic potential contour lines (c
orcho
a
lVaraDo
et alii, 2009; z
inn
& K
oniKow
, 2007; S
chöttler
,1997,
Fig. 2). commonly a downward component of the groundwater
flow exists in recharge areas, and an upward component in dis-
charge areas. Vertical flows occur within long-screen wells, even
in homogeneous aquifers with very small vertical head differences
(B
aSiricò
et alii, 2015).
in general, groundwater flow focuses on layers and structures
of high hydraulic conductivity resulting in different hydraulic
pressures within the aquife
r. this situation can lead to aberrant
piezometric heads measured
in wells that access layers of differ-
ent hydraulic pressure. thus, water table contour lines based on
piezometric well head measurements can show deformations that
are caused by these heterogeneities.
P
hilliPS
(1991) analytically described the role of permeable
lenses on the flow network in some simple cases. F
rancani
et
alii (2002) applied similar relations to evaluate the flow and pi-
ezometric head distribution near lenses of different hydraulic
conductivity. numerical models can describe in 3D the effect of
heterogeneities on piezometric head distribution and flow lines
(e
aton
, 2006; g
attinoni
, 2011).
nevertheless, in real cases the
substantial issue is to locate the significant heterogeneities in
reference to the studied problem and characterize them, from a
geometrical and hydrogeological point of view, to formulate a
reliable conceptual model of the aquifer with the suitable level of
detail. the possibility to identify and to characterize the hydraulic
impact of individual lenses and layers on the local flow network
depends on the numbers, density and quality of available data.
in recent studies, in-hole vertical groundwater measures were
used to support the formulation of the groundwater flow conceptual
model (M
aStrocicco
et alii, 2013; P
etitta
et alii, 2013).
in order to support conventional methods of aquifer
characterization, two direct measuring procedures are proposed
and their application discussed in a case study. Vertical and
horizontal groundwater velocity and direction are measured
in boreholes (free of using artificial tracers) to display the
heterogeneity of the aquifer and to evaluate the role of layers
with lower hydraulic conductivity. coupled to a piezometric
monitoring program, the gained results can help to identify layers
that act as aquitard and help to detect critical hydrogeologic
conditions in regard to a studied problem.
a numerical model has been developed to analyze the
differences between a conceptual model of the aquifer solely based
on conventional methods of characterization and a model that
comprises data from the proposed borehole flow measurements.
Applied methods
the suggested approach is based on in-hole groundwater
flow measurements to check the conceptual model of the aquifer.
in the presented case study the following survey procedure was
proposed and executed:
tracer-free measurements of ambient vertical groundwater
background image
ApplicAtion of borehole flow meAsurements to chArActerize hydrAulic heterogeneities And their impAct
on the locAl groundwAter flow network
7
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
flow in monitoring wells that are screened along the whole
aquifer profile. this measurement comprises a flow-log in
undisturbed conditions along the screened well section in
order to identify hydraulic short cuts and to localize sections
of ambient inflow and outflow along the well profile. the
aim is to assign different piezometric pressures on the
corresponding strata and to identify layers that might act as
aquitards. Subsequently, this measurement procedure was
repeated under forced flow conditions (low-rate pumping
directly below the groundwater level) in order to quantify
and assign a hydraulic conductivity profile to the surrounding
strata.
Measurement of piezometric heads in cluster piezometers
filtered at depth corresponding with the detected main inflows
and outflow sections.
tracer-free horizontal in-hole flow velocity measurements
in the depth of strata with high hydraulic potential, both in
piezometers with long screens and in cluster piezometers
with short screens. Measurement taken in piezometers with
short screen intervals are more representative, as the natural
hydraulic situation of the individual layer is less disturbed.
Vertical groundwater velocity measurements
if aquifer layers with different pressure potentials are
connected by a well bore, a hydraulic short circuit is induced in the
well and a permanent vertical groundwater flow in the well bore is
the result. the flow velocity varies along the well profile in regard
Fig. 1 -
Example of hydrogeological sections describing the alluvial aquifer of the Milan area in East-West (OE) and North-South (NS) extension, cor-
responding with the maximum slope (modified from D
enti
et alii, 1993)
background image
V. frAncAni, p. gAttinoni, m. schöttler, s. mottini, d.s. rigAmonti & m. Vitiello
8
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
to the hydraulic potential of individual horizons. to detect the
presence of a hydraulic short-circuit, to quantify the vertical flow
and to characterize the variability of hydraulic conductivity along
the aquifer profile, a thermo-flowmeter was used. the Berghof
thermo-Flowmeter is capable of sensing very small flow motion
in water (down to less than 1 mm/s) by the principle of constant
temperature anemometry (cta) measurements (h
alla
, 2006). it
is similar to a Fluid-log with the advantage that no artificial tracers
are needed. the working principle is based on the cooling effect
of a flow on a heated body. the cta measures velocity along the
axial well profile and provides continuous velocity readings that
are processed into amplitude and time-domain statistics.
thermo-flowmeter measurements were carried out both
under natural undisturbed hydraulic conditions and under forced
hydraulic conditions (low-rate pumping directly below the
groundwater table). in the former, hydraulic short circuits are
detected and inflows/outflow horizons are identified. in the latter,
groundwater inflow is forced along the well screen profile: the
inflow quantity from the different horizons corresponds with
their hydraulic conductivity, that can be quantified by the inflow/
withdrawal ratio in the well profile. the precondition for this
calculation is that the average hydraulic conductivity over the
whole aquifer profile is known (i.e. from pumping tests).
Horizontal groundwater velocity measurements
horizontal groundwater flow velocity measurements were
carried out with the Phrealog measuring system. this technology
is based on the observation and optical recording of the movement
of natural microscopic particles that are present in the groundwater
and carried as suspension along with the natural groundwater flow.
the measurements are carried out in selected intervals in the well
bore. the in-hole flow rate can be tracked within a velocity range
from 1x10
-2
m/s (0.01 m/s) to lower than 1x10
-6
m/s (0.08 m/d).
Depending on lithological conditions, velocities lower than 6x10
-7
m/s (0.05 m/d) are generally set to zero or to “no flow movement”.
naturally occurring particles are carried along by the in-hole
flow that corresponds with the ground water flow system in the
aquifer (S
chöttler
, 2004). the observed particles often consist
of aggregations of i.e. organic and inorganic material as clay
minerals. at high flow rates the transportation of silt particles and
even fine sand fractions can be observed.
in order to measure groundwater flow direction and velocity,
images of a illuminated focal plane in the axial centre of the
measuring section are continuously recorded by camera. the
image sequences are computed in real time by a parti
cle tracking
software calculating the drift of the particle patterns. the flow
direction is determined with the probe-integrated compass
and the flow velocity is calculated by timing the movement of
particle patterns across the cameras field of view. to calculate
the Darcy Velocity based on the measured in-hole flow, two
correction factors are used: the factor α and the factor γ
(o
gliVi
,
1958;
K
rätzSchMar
& l
ucKner
, 1966; K
lotz
, 1977, 1978). the
product between α and γ is equal to V
H
/V
f
, where V
H
is is the flow
velocity along the median flow line within the well, and V
f
is the
Darcy velocity.
the factor α compensates the influence of the flow
geometry caused by well construction. this factor is calculated
by the hydraulic conductivity K of the aquifer, the filter pack and
filter screen as well as the diameter of the bore and the well. the
factor γ considers the influence of the probe on the in-hole flow
(see the following references for the complete formula of these
correction parameters: D
roSt
et alii, 1968; S
chöttler
, 1997;
2004; 2007). Furthermore, the hydraulic influence of the well
geometry in the investigated sections can be simulated and the
data aligned with a respective simulated flow scenario (D
rieSSen
et alii, 2015). Simultaneous to flow tracking, the particle load is
quantified. this additional information can be used to support the
identification and estimation of possible piping effects at an early
stage as it can help to prove and quantify the outwash/relocation
of fine grain material within the aquifer.
Fig. 2 -
Extreme heightened proportions of a potential flow field dis-
turbed by a fully screened well (S
chöttler
, 1997, modified after
B
ergmann
, 1970)
background image
ApplicAtion of borehole flow meAsurements to chArActerize hydrAulic heterogeneities And their impAct
on the locAl groundwAter flow network
9
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
cAse study
Background and framework description
the discussed survey procedure was applied to a case study;
the location and site details are undisclosed in order to meet
privacy protection requirements.
an underground structure had to be realized at the bottom
of a hill slope, in a heterogeneous aquifer, with the following
schematic stratigraphy:
from ground surface to 10 m depth: gravel to sandy silt;
from 10 m to 28/40 m depth: fine-medium sand and silty
sand to medium-coarse sand, from loose to moderately packed;
from 28-40 m depth: rock, from friable to compact (slightly
consolidated), 2-3 mm joints filled with calcium carbonate.
the hydrogeological setting of the area is characterized by a
multilayer aquifer having a groundwater flow directed north to
South (Fig. 3).
in the first phase of characterization, a pumping test was
conducted to quantify the transmissivity along the aquifer profile
and a mean overall transmissivity T of 0.04 m
2
/s was calculated.
Moreover, lugeon tests were available referred to the rock
formation at the bottom of the hydrogeologic system, whose results
show a permeability range from 2 x 10
-7
m/s to 3 x 10
-6
m/s. as a
consequence, the rock formation is considered as a bedrock, with
a hydraulic conductivity that can vary locally, depending on the
degree of fracturing and alteration of the rocks.
a supplementary hydrogeologic survey was conducted
to improve the existing conceptual model of the aquifer in the
vicinity of the underground construction.
Results
in order to verify the on-site hydraulic situation and to
determine ground water flow direction and velocities, in-situ
vertical and horizontal ground water flow measurements were
initially conducted in four monitoring wells
realized
with long
filter screen intervals of more than 25 m (lF1, lF2, lF3 and
lF4, Fig. 3).
With regard to the heterogeneous lithology of the aquifer,
groundwater head variations along the vertical aquifer profile
were assumed. consequently, thermo flowmeter logs (vertical
velocity measurements) were conducted:
1. under the normal, undisturbed conditions, in order to detect
Fig. 3 -
Monitoring wells with long filter screens (LF) and cluster monitoring wells (CL) available on site. Structure A is bound by diaphragm walls,
driven down into the bedrock to 40-45 m. Structure B is an underground structure located at aquifer depht from 18 to 27 m b.g.s., which may be
considered as an impermeable barrier to underground flow. The depicted piezometric lines are inferred by bibliographic studies on a large scale,
previous to underground working activities
background image
V. frAncAni, p. gAttinoni, m. schöttler, s. mottini, d.s. rigAmonti & m. Vitiello
10
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
short-cuts (as effect of different hydraulic heads in the
layers), depth and productivity of inflow horizons along the
well profile
2. while pumping with a low rate, in order to quantify the
hydraulic conductivity profile of the aquifer (the pump was
positioned directly below ground water level).
the profiles of the thermo-Flowmeter logs conducted under
undisturbed hydraulic conditions revealed significant upward
flow in the wells, indicating a hydraulic short-cut from the lower
into the upper part of the aquifer, and revealed confined hydraulic
conditions in the lower section of the aquifer. the same situation
was registered in all four wells in similar depth sections: from
around 20 m b.g.l. and down to 33 m b.g.l. the aquifer in the area
of piezometers lF1-lF4 is characterized by several horizons with
confined groundwater and with high hydraulic conductivity. these
horizons vary in thickness from 0.5 m up to 4 m. the piezometric
head in the depth section from approximately 20 m down to 36 m
b.g.l. is higher than the piezometric head in the overlying layers,
thus resulting in an upward infiltration from the lower into the upper
aquifer and an upward flow in the wells via hydraulic short-cuts.
the profiles of thermo-flowmeter logs conducted under forced
hydraulic conditions (while pumping) in all the four monitoring
wells revealed a sequence of high and low permeable horizons,
varying from 1.7 x 10
-5
m/s (lF1, 0-8 m b.g.s) to 3.3 x 10
-2
m/s
(lF3, 9-9.3 m b.g.s).
as upward flow was recorded in all four proximate wells,
a intermediate layer sequence acting as an aquitard must be
assumed for this site section. this situation can be characterized
as a “leaky aquifer”, in the sense of an upward directed leakage.
Moreover the thermo-flowmeter logs identified the presence of
layers with high hydraulic conductivity in the bottom section of
the aquifer that act as main discharge horizons.
Subsequently, new cluster monitoring wells were realized
(named cl in Fig. 3), with short screen sections of 3 to 6 m length,
to avoid hydraulic short-cuts, placed at depth corresponding
with the detected main inflow and outflow horizons. the new
monitoring points confirmed higher piezometric heads for
the lower part of the aquifer (in particular cluster cl7, cl6,
cl2, placed upstream of the underground structures, Fig. 4),
locally perturbed or amplified by the presence of impermeable
underground structures in the upper layers of the aquifer (cluster
cl1, cl5, cl4, cl3, Fig. 4).
to quantify the flow in the more productive layer detected at
the bottom of the aquifer, horizontal flow measurements in lF1,
lF2, lF3 and lF4 were performed in sections approximately
from 25 m b.g.l. down to 34 m b.g.l.. the order of magnitude of
the calculated mean horizontal flow velocity (Darcy velocity) in
wells upstream of the underground structure B (lF1 and lF3) is
10
-4
m/s (tab. 1). these monitoring wells connect all the aquifer
layers and the upward gradient monitored on site produces
upward flows in the wells. in spite of the hydraulic isolation of
the measuring cell with packers, for some depths the enhanced
velocity caused by vertical flow in the vicinity of the well (skin
effect) affects the horizontal velocity flow measurements.
in the new cluster piezometers with short filter screens (cl,
tab. 1), the horizontal in-hole flow measurements are not affected
by upward flow and the results are more representative to display
the undisturbed situation of the aquifer. For example, cl2-1 taps
the lower part of the aquifer from 29 m to 33 m b.g.l. and the
calculated mean Darcy-velocity over all measured depth is 1.4 x
10
-5
m/s, significantly lower than documented for the near well
lF3 (3.3 x 10
-4
m/s).
the horizontal velocity measures conducted in short screened
monitoring wells show that in the deeper aquifer layers the
steady horizontal flow can generally be assumed as low (order
of magnitude 10
-5
m/s to 10
-6
m/s) under undisturbed natural
conditions. higher velocities have been monitored in clusters
cl3, cl4 and cl5, placed downstream to the structure B, that
is an impermeable barrier from 18 to 27 m b.g.s. this result
indicates enhanced groundwater flow below structure B, as the
permeability of the underlying layers allow a higher groundwater
passage than it was formerly given by the natural underground
structure of the aquifer without the barrier function of structure B.
Discussion
a simple numerical model of groundwater flow was developed
in order to compare two conceptual models of the aquifer: one
obtained by conventional methods of characterization and the other
obtained by data from the new proposed survey. in particular, two
model configurations were used in this study. the first, referred
as scenario 1, was developed to describe the conceptual model
obtained by conventional methods of characterization (meaning,
in this case, stratigraphic logs, pumping tests and lugeon tests).
the second (scenario 2) updated the parameter value (hydraulic
Tab. 1 -
Mean horizontal Darcy velocity recorded in monitoring
background image
ApplicAtion of borehole flow meAsurements to chArActerize hydrAulic heterogeneities And their impAct
on the locAl groundwAter flow network
11
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
Fig. 4 - P
iezometric heads (m. a.s.l.) monitored in cluster piezometers for T1 (February 14
th
), T2 (February 15
th
) and T3 (March 7
th
) . In brackets: filters
screen depth below ground level. The position of cluster piezometers is shown in Fig. 3
background image
V. frAncAni, p. gAttinoni, m. schöttler, s. mottini, d.s. rigAmonti & m. Vitiello
12
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
conductivity) in the base model to match field data collected
during flowmeter surveys.
the USGS computer program MoDFloW (h
arBaugh
et alii, 2000; h
arBaugh
, 2005) was used for this analysis. the
model was not detailed enough to be fully calibrated by on site-
data, but it was useful to analyze the major differences in the
conceptual models of the aquifer.
the finite-difference grid for the numerical model consisted
of uniformly spaced model cells that were 5 m on one side. the
grid consisted of 88 rows, 60 columns and 12 layers that covered
an area of around 0.13 km
2
(0.3 x 0.44 km).
the following boundary conditions were set: no recharge
via ground surface and constant heads (ch) at the upstream and
downstream sides of the system representing the piezometric
level monitored in available piezometers (1.9 m a.s.l at the north
side and 0.3 m a.s.l at South). the ground surface was reproduced
according to the data obtained by stratigraphic logs and
monitoring points available on site. the layers were not sloping
between the South end of the studied area and the north end of
the underground structure, while they had a slope equal to the one
of the ground level between the north end of the underground
structure and the upstream end of the model. as a consequence
of the morphological characteristics of the area, the north ch
condition was placed only in saturated layers.
in scenario 1 the aquifer was represented as an equivalent
porous media with uniform properties: hydraulic conductivity
value of all layers corresponding with incoherent aquifer
materials (tab. 2 and Fig. 5, K
x
=K
y
, layers 1-10) was constant
and equal to the K obtained by the pumping test (1.1 x 10
-3
m/s).
hydraulic conductivity values of the two layers at the bottom of
the model were equal to 1 x 10
-6
m/s and 1 x 10
-7
m/s, represented
the bedrock, and
they
were differentiated according to the results
of lugeon tests. the vertical hydraulic conductivity of layers (K
z
)
was always equal to 1/10 of K
x
.
in Fig. 5, the flow net along a section was represented; the
more superficial layers of the model were unsaturated at the north
side; the piezometric lines showed some deformations in layers
corresponding with the bedrock but
did not show vertical gradient
in layers corresponding with incoherent aquifer materials.
Subsequently, the numerical model was modified to represent
the changes in the conceptual model of the aquifer introduced by
the results of the thermo-flowmeter logs.
in scenario 2 the layers represented the heterogeneities of
individual strata, each with different hydraulic proprieties: data
measured by thermo-flow meter log of lF3 were used to modify
the hydraulic conductivity values of layers corresponding with
the incoherent aquifer materials (tab. 2 and Fig. 6, K
x
=K
y
, layers
1 to 10).
this approach includes the assumption of extension of the
hydraulic conductivity measured at lF3 monitoring well to the
whole domain of the model, with the aim of evaluating how
such variability of K values could modify the hydraulic net-
work. the two layers at the bottom of the model had the same
hydraulic conductivity as the previous model and the vertical
hydraulic conductivity of layers (K
z
) was always equal to 1/10
of K
x
. in Fig. 6 the flow net along the same section as Fig. 5 was
represented and showed significant upward gradient in strata
corresponding with incoherent aquifer materials. this change
was due to the presence of strata with lower hydraulic conduc-
tivity (for example 6.1 x 10
-5
m/s from 18 to 23 m depth) that act
as an aquitard. the upward gradient caused the upward flows
detected on-site in monitoring wells with long filter screen that
connect all the aquifer layers.
the results of the study modified the conceptual model of
the aquifer and illustrated a potentially critical hydrogeological
situation: if the lower, partially confined layers of the aquifer
are intercepted by structures that modify the local hydraulic
situation, the high permeability of these horizons could lead
to an increase of the groundwater flow velocity to a degree
that might not be compatible with the technical parameter of
soils. For example, D
ella
r
oSSa
et alii (2003), g
attinoni
&
F
rancani
(2009) stated that the overcoming of the critical speed
of groundwater flow introduces erosion effects that can lead to a
destabilization
of soil structures.
the proposed methodical approach can be applied for site
investigations prior to, or in the course of, ongoing underground
construction activities in order to identify hydrogeologic
circumstances that can be critical to construction works in aquifers.
Tab. 2 -
Comparison between the aquifer hydraulic conductivity
scheme according to conventional methods of characterization
(scenario 1) and data from the new proposed approach (sce-
nario 2, k data referred to LF3)
umulative curve showing the
grain size composition of the soil sampled in the excavation for
the permeability test in situ
background image
ApplicAtion of borehole flow meAsurements to chArActerize hydrAulic heterogeneities And their impAct
on the locAl groundwAter flow network
13
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
main differences in the conceptual model of the aquifer obtained
by conventional methods of characterization and in the updated
one obtained through the new approach of investigation. the first
was characterized by a regular pattern of flow, without vertical
gradients in layers corresponding with incoherent aquifer materials.
the second, instead, showed significant upward gradients in strata
corresponding with incoherent aquifer materials. this change is
due to the presence of strata with lower hydraulic conductivity that
act as an aquitard and locally confine the more permeable strata
detected at the bottom of the aquifer. even if the numerical model
is not fully calibrated by site data, it supports the field
study to
understand and describe the complex hydrogeological
situation.
the survey method proposed in this study leads to accurate
results on a local scale, whose significance in the context of the
studied issue must be discussed by relating the collected data
with other information on the hydrogeological system. Vertical
and horizontal groundwater velocity measurements in monitoring
points designed ad hoc, can substantially change the conceptual
model of the aquifer; they can be integrated with the conventional
techniques of characterization of the aquifer, assuming the
following operational steps:
1. characterization of soils with hydraulic conductivity tests,
piezometric head monitoring and in-hole groundwater flow
velocity measurements;
2. processing and merging of collected data with the elaboration of a
conceptual model of the aquifer (the evaluation of groundwater
conclusions
the subject of this paper is an advanced investigation approach
to identify and characterize significant hydraulic heterogeneities
in an aquifer, with a suitable level of detail in reference to a
studied case. conventional hydrogeologic characterization is
generally based on data from stratigraphic logs, pumping tests
and piezometric monitoring. these data, applied on a local scale,
are in some cases not sufficient to resolve the hydraulic situation
with an appropriate level of detail, especially in the vertical
profile of the aquifer. in order to support conventional methods
of characterization, two direct in-hole vertical and horizontal
groundwater flow velocity measuring procedures are proposed.
they were applied in a case study where supplementary
hydrogeologic surveys were conducted to improve the existing
conceptual model of the aquifer near an underground structure.
the additionally provided data completed the information
available for modeling the aquifer with sufficient accuracy.
Vertical thermo-flowmeter logs allowed the detection of some
noticeable heterogeneities: a leaky aquifer, an upward directed
leakage, and several layers with higher hydraulic conductivity in
the lower part of a hydrostratigraphic unit that appeared almost
homogeneous. Moreover, horizontal in-hole groundwater flow
measurements allowed for quantifying the flow in the more
productive layers with observed Darcy velocity up to 10
-4
m/s in
some measuring points.
a numerical model was applied in two scenarios, to analyze the
Fig. 5 -
Piezometric lines (side view, NS section) and flow directions
resulting from the conceptual model of the aquifer obtained in
scenario 1 (a = incoherent aquifer materials; b = bedrock)
Fig. 6 -
Piezometric lines (side view, NS section) and flow directions
resulting from the conceptual model of the aquifer obtained in
scenario 2 (a = incoherent aquifer materials; b = bedrock)
background image
V. frAncAni, p. gAttinoni, m. schöttler, s. mottini, d.s. rigAmonti & m. Vitiello
14
Italian Journal of Engineering Geology and Environment, 2 (2015)
© Sapienza Università Editrice
www.ijege.uniroma1.it
these stages should be completed by validation of numerical
model results, achieved by their comparison with real piezometric
and flow variations, for example induced by wells or pilot
excavation.
velocity achievable by indirect approaches cannot reach the
level of detail of velocity measurements in boreholes);
3. numerical modeling of natural flow and forecast of its variations
caused by the planned underground activities.
Received September 2015 - Accepted November 2015
references
B
aSiricò
S., c
roSta
g.B., F
rattini
P., V
illa
a. & g
oDio
a. (2015) - Borehole flowmeter logging for the accurate design and analysis of tracer tests. Ground-
water theme issue 2015, 53-S1: 3-9.
B
ergMann
h. (1970) - Über die Grundwasserbewegung am Filterrohr. GSF-report r24, Gesellschaft für Strahlenforschung München-neuherberg.
c
orcho
a
lVaraDo
J.a., B
arBecot
F. & P
urtSchert
r. (2009) - Ambient vertical flow in long-screen wells: a case study in the Fontainebleau Sands Aquifer
(France). hydrogeology Journal, 17: 425-431.
D
ella
r
oSSa
M.c., F
rancani
V. & g
attinoni
P. (2003) - Studio idrogeologico del territorio monzese: individuazione e caratterizzazione delle zone a bassa
resistenza. Quaderni di geologia applicata 10-2.
D
enti
e., F
rancani
V., r
inelli
S. & S
ala
P. (1993) - Criteri idrogeologici per l’ottimizzazione dell’attività estrattiva nella Provincia di Milano in funzione
della compatibilità ambientale. Quaderni di tecniche di protezione ambientale, 27, Pitagora editrice, Bologna.
D
rieSSen
J., S
chöttler
M., e
nzMann
F., l
aKDawala
z., S
teiner
K., P
oPoV
P., o
leg
i., D
rewS
M., w
ieBer
g. & K
erSten
M. (2015) - PHREASIM - Ein
Expertensystem zur Simulation von Fließverhältnissen in Grundwassermessstellen und deren unmittelbarem Nahfeld. Grundwasser, 20 (3): 181-195.
D
roSt
w., K
lotz
D., K
och
h., M
oSer
h., n
euMaier
F. & r
auert
w. (1968) - Point dilution method measuring ground water flow by means of radioisotopes.
Water resources research 4, 125-147.
e
aton
t.t. (2006) - On the importance of geological heterogeneity for flow simulation. Sedimentary Geology, 184: 187-201.
F
rancani
V., g
attinoni
P., l
ongoni
l. & M
olteni
a. (2002)- Circolazione idrica in strutture idrogeologiche complesse. iGea, 17: 3-15.
g
attinoni
P. (2011) - Connectivity of an alluvial aquifer: from theory to practice. in recent advances in environment, energy Systems and naval Science,
Proceeding of the WSeaS int conf, Barcellona, Spain September 15-17: 89-92.
g
attinoni
P. & F
rancani
V. (2009) - A tool for modeling slope instability triggered by piping. World academy of Science, engineering and technology, 3.
h
alla
P. (2006) - Messung vertikaler Durchlässigkeitsverteilungen mittels Thermoflow. Proceedings Symposium “Vor-ort-analytik - Feldmesstechnik für
die erkundung von kontaminierten Standorten”, 28.-29.11.2006, Stuttgart, Germany
h
arBaugh
a.w. ( 2005) - MODFLOW-2005, The U.S. Geological Survey modular ground-water model. The ground-water flow process. U.S. Geological
Survey techniques and Methods 6-a16, variously pp.
h
arBaugh
a.w., B
anta
e.r., h
ill
M.c. & M
c
D
onalD
M.g. (2000) - MODFLOW-2000, The U.S. Geological Survey modular ground-water model - User
guide to modularization concepts and the ground-water flow process. U.S. Geological Survey open-File report 00-92, 121 pp.
K
lotz
D. (1977) - Berechnung der Filtergeschwindigkeit einer Grundwasserströmung aus Tracerverdünnungsversuchen in einem Filterpegel. GSF-Bericht
r 149; 50 S., 24 abb., 8 tab.; München.
K
lotz
D. (1978) - Alpha-Werte ausgebauter Bohrungen. GSF- Bericht r 176; 119 S., 12 abb., 100 tab.; München.
K
rätzSchMar
h. & l
ucKner
l. (1966) - Analoge elektrische Strömungsuntersuchungen zur Klärung des Zusammenhangs zwischen der messbaren Konzen-
trationsabnahme eines Tracers in einem Pegel und der natürlichen Grundwassergeschwindigkeit. Wasserwirtschaft u. –technik 16: 122-125.
M
aStrocicco
M., S
BarBati
c., c
oloMBani
n. & P
etitta
M. (2013) - Efficiency verification of a horizontal flow barrier via flowmeter tests and multilevel
sampling. hydrological processes, 27: 2414-2421.
o
gliVi
n.a. (1958) - Electrioliceskij metod opredelenija skorostej filtracii. Bjull.o.n.t.i., 4, Gosgeoltehizdat, 42 S., Moskau 1058.
P
etitta
M., P
acioni
e., S
BarBati
c., c
orVatta
g., F
anelli
M. & a
raVena
r. (2013) – Hydrodinamic and isotopic characterization of a site contaminated
by chlorinated solvents: Chienti River Valley, Central Italy. applied geochemistry, 32: 164-174.
P
hilliPS
o.M. (1991) - Flow and reactions in permeable rocks. cambridge University,Press. iSBn0 521 38098 7.
S
chöttler
M. (1997) - Meßbarkeit der Grundwasserbewegung durch Visualisierung der Strömung in Bohrbrunnen-Diss. Univ. Köln; 119 S.; Shaker Verlag,.
S
chöttler
M. (2004) - Erfassung der Grundwasserströmung mittels des GFV-Messsystems. Geotechnik, 27 (1); Deutsche Ges. f. Geotechnik; S.41-45; 8
abb.; Verlag Glückauf, essen.
S
chöttler
M. (2007) - Ein neues Verfahren – Die Grundwasser-Fluss-Visualisierung (GFV). energie | wasser-praxis 12/2007 – DVGW Jahresrevue – hrsg.:
DVGW e.V.; S. 32-37; 6 abb.; wvgw Wirtschafts- und Verlagsges. Gas und Wasser mbh; Bonn.
z
inn
B.a. & K
oniKow
l.F. (2007) - Effects of intraborehole flow on groundwater age distribution. hydrogeology Journal, 15: 633-643.
Statistics