Groundwater Thermal Monitoring to Characterize

Available online at www.sciencedirect.com
ScienceDirect
Procedia Environmental Sciences 25 (2015) 199 – 205
7th Groundwater Symposium of the
International Association for Hydro-Environment Engineering and Research (IAHR)
Groundwater thermal monitoring to characterize streambed
water fluxes of the Brenta river (Northern Italy)
G. Passadorea*, A. Sottanib, L. Altissimoc, M. Puttid and A. Rinaldoa,e
a
Dipartimento di Ingegneria Civile Edile e Ambientale e Centro Internazionale di Idrologia “Dino Tonini”, Università di Padova,
via Marzolo 9, 35131 Padova, Italy
b
Sinergeo S.r.l., Contrà del Pozzetto 4, 36100 Vicenza, Italy
c
Centro Idrico di Novoledo, via Palladio 128, 36030 Villaverla (Vicenza), Italy
d
Dipartimento di Matematica, Università di Padova, via Trieste 63, 35121 Padova, Italy
e
Laboratory of Ecohydrology ECHO/IIE/ENAC, Ecole Polytechnique Fédérale Lausanne (EPFL), Station 2, GR C1 575, 1015
Lausanne, Switzerland
Abstract
The work proposed deals with the characterization of temporal and spatial variability of water exchange fluxes
from/to the Brenta river streambed (Veneto, Italy), critically important to regional water resources management. The
aquifer system evolves from a large undifferentiated aquifer close to the adjacent mountain ranges, later developing
into a well-structured multi-layer system made up of six well-defined confined aquifers constituting the noteworthy
subsurface reservoirs of the area. A three-dimensional groundwater flow model of the multi-aquifer system of the
Central Veneto, within a 3300 km2 area, integrating a large amount of data, has been developed to analyze the
behaviour of this large, complex and well-monitored sedimentary system. The temporal and spatial variability of the
water exchange flux is investigated using heat as a tracer in conjunction with water-level measurements (in the river
and in piezometers close to the river): an effective method for estimating groundwater/surface water exchanges. In
the study area, 5 groundwater monitoring wells (4 on the right bank, 1 on the left bank) were drilled along the river
banks near a pilot transversal ramp (permeable river barrier) built in order to reduce the downstream flooding risk and
to increase the river dispersion. A hydrothermal model of the river and of the underlying aquifer was implemented to
improve the estimation of the water exchange between the Brenta river and the aquifer system. The contribution of
exchange between the river and the underlying aquifer is quantified by comparing simulations of water flow and heat
transport to observed temperature and levels in river and in groundwater.
©
by Elsevier
B.V. This
an open access
© 2015
2015Published
The Authors.
Published
by isElsevier
B.V. article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of the IAHR Groundwater Symposium 2014.
Peer-review under responsibility of the Scientific Committee of the IAHR Groundwater Symposium 2014
Keywords: Groundwater flow, Heat transport, Streambed flux, Geothermal surveys, Temperature, Numerical modelling
* Corresponding author. Tel.: +39 349 827 5445; fax: +39 049 827 5446
E-mail address: [email protected]
1878-0296 © 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of the IAHR Groundwater Symposium 2014
doi:10.1016/j.proenv.2015.04.027
200
G. Passadore et al. / Procedia Environmental Sciences 25 (2015) 199 – 205
1. Introduction
The area of interest is located in North-Eastern Italy (fig. 1) encompassing the large, deep multiaquifer
groundwater reservoir formed in the quaternary de-posits of the Po plain sedimentary basin and the
Brenta river system [1] [2]. In this area several projects are under investigation to withdraw water from
the Brenta river both for aquifer recharge and anthropic consumption. Wise management of this resource
needs to consider potential effects upstream and downstream of the area of interaction of the intimately
coupled river and subsurface system. Spring emergence (termed risorgive, i.e. the spring sites where
spatially widespread surficial emergence of the first phreatic aquifer occurs ephemerally) characterizes a
broad area located roughly in correspondence to the transition from the unconfined to the confined aquifer
system (fig. 1).
Fig.1. Study Area (Central Veneto - Italy), model domain, boundary conditions, and phreatic surface - piezometric head contours
with 10 m steps
Recharge is mainly originated from the fractured system forming the Pre-Alps (the Asiago Plateau), by
rain and by significant streambed dispersion of the Brenta river flowing from the Alps. A distinct feature
of the subsurface basin is the presence of an undifferentiated unconfined aquifer system formed by
alluvial deposits developing on a strip of roughly 5 to 20 km bordering the piedmont region and
characterizing the entire Venetian plain. Outside this strip, a well-defined confined, multiaquifer system
originates from the alternation of silty-clayey layers and sandy fine-grained aquifers. Both the
undifferentiated and the multiaquifer systems are heavily exploited for urban, industrial and agricultural
uses. The depth of the phreatic surface with respect to the ground level starts at a few tens of meters in the
piedmont zone and gradually decreases to zero developing in a strip of spatially distributed plain springs
just south of the appearance of the first low permeability lens, called the area of the risorgive (or
G. Passadore et al. / Procedia Environmental Sciences 25 (2015) 199 – 205
fontanili). All groundwater levels maps realized using several data acquisition campaigns revels a
different behaviour between the upper part of the Brenta river (from Bassano to spring emergence
groundwater flux lines are diverging from the river) and the lower part of the Brenta river (from spring
emergence to Carturo groundwater flux lines converge towards the river). The upper portion of the river,
of about 12 km, is characterized by stream levels greater than groundwater levels and so there is a water
flux from river to aquifer through streambed; in the lower part of the Brenta river, of about 13 km, stream
levels are lower than groundwater levels (which are close to ground level) and the aquifer recharges the
river through streambed. The transition between dispersing and draining streambed depends on
groundwater levels: in the periods characterized by high groundwater levels the inversion of the
behaviour of the river shifts towards the north because the phreatic surface intersect the ground level
further north.
2. Material and Methods
The three-dimensional (3D) flow model of the aquifer system [1] [2] is calibrated through the large
number of available observational data, chiefly concerning phreatic and piezometric surfaces, censuses,
and evidence gathered from pumping tests. The water balance includes the following components: 1) the
net infiltration of the rain, 2) the water discharge dispersed and drained by rivers and 3) by irrigation
channels and fields, 4) water discharges from the springs and 5) the water fluxes springing from, or
pumped out by, a large number of private and public wells operating in the area. The water flux
exchanged between the Brenta river and the underlying aquifer represents an important rate to the
recharge of the aquifer system and it is about the 22% of total ingoing water flux (average annual values
of about 12 m3/s out of 50 m3/s). The three-dimensional geological model of the system (i.e. its
conceptual model), set up by means of extended geological sections and stratigraphies, is used to create
the unstructured finite elements grid. Necessary field data concern the local hydraulic
conductivity/transmissivity of the aquifers. The steady state calibration is carried out based on real
phreatic surface obtained by interpolation of about 100 phreatic observational wells. Estimates of water
exchange have been carried out in the past by means of differential discharge measurements along the
river (fig. 1).
The investigation campaigns of the past have permitted to extract a correlation between the river
discharge and water dispersion through the streambed: if the stream flow discharge is of about 10 m3/s the
water exchange toward the aquifer is of about 7 m3/s, when water discharge of the river is of about 150
m3/s, the flux ingoing to the aquifer is about 20 m3/s. The draining nature of the lower part of the river
(characterized by a water flux from the aquifer to the river) has not been investigated adequately because
of the difficulty to conduct differential discharge measurements in the second portion of the river
characterized by large and changeable sections, by a great number of gains and losses from surface
channels and by the presence of several expansion areas very close to the river and variously
interconnected with it. The drained groundwater discharge from the first unconfined aquifer seems to be
equal or slightly higher than the water flux from the river to the aquifer of the upper part of the river. So
the discharge of the river at the beginning and at the end of the dispersing and draining streambed is
approximately equal, taking into account the uncertainty of these measurements. The accuracy of this
type of measurements is limited by partially known river diversions and uncertainties in discharge
measurements, especially during extreme events. Moreover, they do not give any information on the
temporal and spatial variability of the exchange,
Using heat as a tracer, the simplest and most available natural tracer along the river channel and its
aquifers, allows to investigate the temporal and spatial variability of the water exchange flux. Infact,
using heat as a tracer, in conjunction with water-level measurements (within the river and in nearby
201
202
G. Passadore et al. / Procedia Environmental Sciences 25 (2015) 199 – 205
piezometers) has been shown to be an effective method for estimating groundwater/surface water
exchanges [3] [4] [5] [6] [7] [8] [9] [10] [11]. This method requires continuous monitoring in time of the
water temperature and the levels of the stream jointly with suitable samples of the temperatures and the
piezometric levels of groundwater (if possible, at multiple depths).
In particular, in the study area, 5 groundwater monitoring wells (4 on the right bank: Pz1, Pz1-1, Pz12, Pz1-3, 1 on the left bank: Pz2) were drilled along the river banks near a pilot transversal ramp
(permeable river barrier) built in order to reduce the downstream flooding risk and to increase the river
dispersion; these points have been monitored with automatic head and temperature data loggers since
March 2007 (Pz1, Pz2) or since June 2014 (Pz1-1, Pz1-2, Pz1-3). A river stage data logger has also been
installed near the permeable check dam in April 2010, in order to obtain hydraulic head and temperature
values of the stream (fig. 2).
Then a hydrothermal model has been developed in order to optimize the seepage estimation and to
simulate the thermal trends observed in monitoring wells. In such manner it has been possible to calculate
the flow rate from the river to the aquifer.
Fig.2. Monitoring wells, river data logger and S1 ramp.
3. Results
Field measurement confirm that groundwater temperature and levels are heavily conditioned by the
streambed dispersion, even in furthest monitoring point.
For example, well Pz2 is nearer to the streambed and the observed values reveal a direct relation with
river leakage. The further well Pz1 starts to show the stream seepage influence only in post-operam
condition, after the realization of a ramp, demonstrating a clear variation in the hydrogeological regime.
Infact, after the completion of the permeable barrier (which consist of the enlargement of the leakage
zone upgradient the ramp) persistent rises of water table (of about 4 meters) have been detected in near
monitoring wells (fig. 3). This condition seems to be stable during a long-term monitoring period which
G. Passadore et al. / Procedia Environmental Sciences 25 (2015) 199 – 205
includes all the different phases of the hydrogeological regime. Thermal datasets proved to be useful to
validate the artificial groundwater recharge effectiveness (fig. 4). For example, in monitoring well Pz1,
that was at first barely influenced by the stream leakage because of the greater distance from Brenta river,
a marked head and temperature variation was identified after the realization of ramp S1.
The river influence on temperature and hydraulic head of the monitoring well (Pz1 and Pz2) is
different in winter or in summer (fig. 5). In winter, a river flood instantaneously raises the hydraulic head
of well Pz1 and well Pz2; the water temperature of well Pz2 drops gradually and reaches the minimum
value in 8-10 days (events 1 and 2 in fig. 5) due to the arrival of the cold water; the water temperature of
well Pz1 does not change due to the great distance from the river (during a flood event). In summer, a
river flood instantaneously raises the hydraulic head of wells Pz1 and Pz2 but the impact on temperature
is not evident due to the lower temperature of the river water and of the groundwater, similar to the
temperature of the flood water (events 3, 4 and 5 in fig. 5). The first model simulations estimate a
medium leakage discharge of about 4 m3/s/km. This value is perfectly comparable with the previous field
measurements, confirming the numerical tool validity for river seepage evaluation.
Fig. 3. Groundwater levels and river head (5 years).
203
204
G. Passadore et al. / Procedia Environmental Sciences 25 (2015) 199 – 205
Fig. 4. Water temperature of the monitoring well Pz1 and Pz2 and of the river (5 years)
Fig. 5. Temperature and hydraulic head of the monitoring well Pz1 and Pz2 and of the river in winter and in summer
G. Passadore et al. / Procedia Environmental Sciences 25 (2015) 199 – 205
References
[1] Passadore G (2008) Modello matematico di flusso nei sistemi acquiferi del Veneto Centrale. Ph.D. dissertation, Università di
Padova
[2] Passadore G, Monego M, Altissimo L, Sottani A, Putti M, Rinaldo A (2012) Alternative conceptual models and the robustness
of groundwater management scenarios in the multi-aquifer system of the Central Veneto Basin. Hydrogeology Journal: Volume
20, Issue 3 (2012), 419-433, doi:10.1007/s10040-011-0818-y
[3] Anderson MP (2005) Heat as a ground water tracer. Groundwater 43 (6), 951–968, doi: 10.1111/j.1745-6584.2005.00052.x.
[4] Constantz J (2008) Heat as a tracer to determine streambed water exchanges, Water Resources Research, 44, W00D10,
doi:10.1029/2008WR006996.
[5] Duque C, Calvache ML, Engesgaard P (2010) Investigating river–aquifer relations using water temperature in an anthropized
environment, (Motril-Salobreña aquifer), Journal of Hydrology 381, 121–133, doi:10.1016/j.jhydrol.2009.11.032.
[6] Engeler I, Hendricks Franssen HJ, Müller R, Stauffer F (2010) The importance of coupled modelling of variably saturated
groundwater flow-heat transport for assessing river–aquifer interactions, Journal of Hydrology 397, 295–305,
doi:10.1016/j.jhydrol.2010.12.007.
[7] Essaid HI, Zamora CM, McCarthy KA, Vogel JR, Wilson JT (2008) Using Heat to Characterize Streambed Water Flux
Variability in Four Stream Reaches, J. Environ. Qual. 37:1010–1023, doi:10.2134/jeq2006.0448.
[8] Kulongoski JT, Izbicki JA (2008) Simulation of Fluid, Heat Transport to Estimate Desert Stream Infiltration, Groundwater, doi:
10.1111/j.1745-6584.2007.00403.x.
[9] Rau GC, Andersen MS, McCallum AM, Acworth RI (2010) Analytical methods that use natural heat as a tracer to quantify
surface water-groundwater exchange, evaluated using field temperature records. Hydrogeology Journal, Volume 18, pages
1093-1110.
[10] USGS (2003) Heat as a tool for studying the movement of groundwater near streams. Circular 1260. U.S. Dep. of the Interior,
U.S. Geological Survey.
[11] Vettorello L, Berti M, Pedron R, Sottani A (2012) Groundwater field measurements for stream seepage estimation. Flowpath
2012, Bologna (Italy)
205