Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,

Architecture and Art – Peila, Viggiani & Celestino (Eds)

© 2019 Taylor & Francis Group, London, ISBN 978-1-138-38865-9

Monitoring the water resources in the large railway projects: The

Fortezza – Ponte Gardena case study, Southern Brenner Base
Tunnel access

S. Rodani, G. Penna & F. Marchese

Italferr SpA, Rome, Italy

A. Scuri, F. Sciascia & D. Castioni

ENSER srl, Engineering company, Faenza (RA), Italy

L. Piccinini
Dipartimento di Geoscienze, University of Padova, Padova, Italy.

F. Cervi
DICAM, University of Bologna, Bologna, Italy

G. Benedetti
Enser s.r.l., Engineering company, Faenza (RA), Italy
DICAM, University of Bologna, Bologna, Italy

ABSTRACT: The Brenner Base Tunnel (BBT) is being built to connect Italy and Austria
running from Fortezza to Innsbruck (55km). The southern access will link Ponte Gardena
with Fortezza. As widely known, underground tunneling works may have significant impact
on regional hydrogeological systems, causing a drawdown of water levels and a decrease of
springs’ discharge rates. Therefore, an extended monitoring of water resources can be crucial
to define the possible interference between the tunneling and the surrounding environment.
The analysis presented in this paper allowed to determine the main hydrogeological regime,
characterizing this portion of the Isarco valley. The hydro-chemical analysis improved the
knowledge of groundwater paths, mainly characterized by low-mineralized flows, running
within the shallow portion of the rock mass or in correspondence of quaternary glacial
deposits and not related to the deep-water network. Finally, isotopic correlations between
rain and underground water have confirmed this hydrogeological layout.

1 INTRODUCTION

The case study presented in this paper refers to the southern access of the Brenner Base
Tunnel (BBT), which is located in Bolzano (northern Italy). The designed tunnels will link
Ponte Gardena to Fortezza (southern entrance of BBT) through two tunnels named Gardena
and Scaleres, 6 km and 15 km long respectively.
The geological layout of the area is characterized by the presence of a metamorphic base-
ment prevalently formed by phyllites with the occurrence of localized Permian granitic and
dioritic intrusions. Thick quaternary deposits (fluvio-glacial and alluvial deposits) usually
cover the slopes and the main valley in which river Isarco flows.
A comprehensive hydrogeological survey started in 2003 and is still ongoing. It consists of
water sampling activities throughout an area larger than the one that will be involved in the
tunnel excavation. This was necessary to identify areas in the vicinity of the tunnel that may be

532
Table 1. Water points surveyed in the area.
Springs Wells Rivers and river diversions Piezometers Rain water collectors

667 183 337 105 6

impacted by tunneling. As a result, a total of 1298 water points (namely: springs, wells, rivers,
piezometers, precipitations) have been monitored in 400 Km2. Water points resumed in Table 1.
Most of water points are represented by springs that can be considered as the most repre-
sentatives of the hydrogeologic network.

2 SPRINGS DISCHARGE RATES

A preliminary analysis of springs’ discharge rates (for 418 springs with at least 1 measure
derived from different data sources such as Provincia Autonoma di Bolzano and Italferr data-
base) have been performed.
The results are shown in Figure 1. They show that springs with average discharge rates
lower than 1 l/s are about the 85%. This means that groundwater network is characterized by
diffuse shallow water flows, which are not able to produce significant water discharge points
(except for some localized area developed along tectonic alignments, such as Scaleres and Spe-
lonca valleys).

3 MONITORING PROGRAM

A reduced number of water points were selected for further hydrochemical monitoring
because of relevant flow rates magnitude and proximity to the tunnel (Table 2).
A total number of 198 points have been monitored in last 2017 campaign.
In Figure 2 the distribution of monitored water points is shown in relation to the railway
track alignment and drainage basin subdivision, useful to determine the presence of water
points clusters with different chemical characteristics and origin.
To define the main hydrochemical characteristics of water points a complete set of ana-
lysis, consisting of physico-chemical parameters (discharge rate, water depth, temperature,

Figure 1. Average discharge rates for springs of Fortezza-Ponte Gardena tunnel.

533
 Table 2. Water points monitored in 2017/18 campaign.
Springs Wells Rivers Piezometers Rain water collectors

96 14 15 67 6

Figure 2. Monitored water points along Fortezza-Ponte Gardena railway track.

534
conductivity, pH, oxygen, ORP and turbidity), major chemical elements (bicarbonate, cal-
cium, chloride. magnesium, nitrate, potassium, silica, sodium and sulphate), trace elements
and water isotopes (δ18O, δ2H, 3H), has been performed.

4 PHYSICAL AND CHEMICAL PARAMETERS OF SPRING WATER

4.1 Specific Conductivity(25°C) SC

This parameter expresses the quantity of dissolved mineral in water and provides useful infor-
mation about the degree of mineralization.
In fact, the analysis is useful to differentiate potential shallow water flows (lower values of
SC) from those of deeper origin (higher values of SC). Based on the homogeneous lithologic
characteristics of the project area (manly phyllites, granites, diorites) the investigation can be
considered reliable for the scope of work.
Data presented in Figure 3 show that about 55% of spring water have a SC lower than
200 μS/cm indicating an overall low degree of mineralization. Anyway, an increase in SC
develops from the top of the mountains (right part of Figure 3) to the valley (left part of
Figure 3) following an inverse exponential relationship. This means that shallow waters pro-
gressively increase their mineralization as a result of the interaction with the rock bulk.
A reduced number of points present unusually higher or lower SC values that deviate from
the exponential relationship. These points have been reported with their identification code in
Figure 3 and further analyzed with water isotopes to estimate their recharge areas (see 6.1).

4.2 TEMPERATURE

A similar relationship than that of SC has been found for water temperature. In fact, springs
at lower altitude exhibit high average temperatures (up to 16°C) while points located at
higher elevation have low average temperatures (up to 4°C). A linear relationship
between springs temperatures and altitudes have been found (Figure 4). As in the case
of SC, monitored water points deviating from the linear relationship have been reported
with their identification code and further analyzed with water isotopes to estimate their
recharge areas (see 6.1).

Figure 3. Specific Conductivity vs Elevation ratio for monitored spring point in study area.

535
Figure 4. Temperature vs elevation ratio for monitored springs in project area.

5 GEOCHEMICAL FACIES

To determine the geochemical characteristics of water a Piper plot diagram is used (Figure 5).
It is a ternary diagram where cations and anions are plotted to define the chemical characteris-
tics of water.
Most of the spring water belong to the Calcium-Bicarbonate hydro-facies (78%). These
waters are the results of dissolution processes taking place within quaternary deposits that
cover the mountain slopes. Even very low quantities of calcite, due to its high solubility, deter-
mine a significant increase in bicarbonate content.
The other main hydro-facies is that of calcium-sulfate water (22%). The sulfate ion is due to
dissolution processes involving Sulphur mineralization within phyllite rocks.
Due to the high number of water samples (around 400), Varna Sud and Bressanone drain-
age basin are reported separately (Figure 6). As represented in Figure 7, well and river water
follow an analogous distribution, in terms of geochemical facies, as springs, confirming the
similar origin of the water flows.

Figure 5. Piper diagram example with different geochemical facies obtained in final diamond plot.

536
Figure 6. Piper plot for springs in Varna Sud and Bressanone drainage basins.

Figure 7. Piper plot for Wells and Rivers across all the area.

Figure 8. Piper plot for Sodium-Bicarbonate water from some of deep piezometers water.

537
 Geochemical analysis of piezometers’ water highlights a distribution like that already seen
for springs. The only difference is represented by the presence of 5 water points (Figure 8) fall-
ing into the sodium-bicarbonate facies. Higher contents in Na+ and/or K+, coupled with high
HCO3 and variable SO42- concentrations is likely attributable to dissolution of alkali minerals
such as aluminosilicates (sodium and potassium plagioclase), supported by the conversion of
aqueous CO2 into bicarbonate ion and, partially, by oxidative dissolution of pyrite. These
samples can be considered as the expression of a deeper and older water circulation.

6 ISOTOPES

Water isotopes provide significant information concerning groundwater flow paths, with
emphasis to recharge altitudes and residence time within the aquifer.
Four water collectors distributed in southern and northern portion of the area (as shown in
Figure 2) have been used starting from 2014 and have been then supported by two additional
PALMEX rain samplers, specifically built for this scope.
This was necessary to compare the isotopic characteristics of rain water with those of
groundwater and to define the variations, in isotopic content, occurred from infiltration to
discharge at single monitored water point.
Analyzed isotopes are represented by 18O (oxygen), 2H (deuterium) and 3H (tritium). The
first two isotopes are stable isotopes, which remain unchanged after their formation (rain
drops), while the third one (tritium) is an unstable isotope that undergoes a radioactive decay
which generates a new chemical element 3He (Helium) and a decay product β-.

6.1 STABLE ISOTOPES (δ 18O, δ2H)

The concentration of these isotopes is expressed as a relative difference δ (‰) between the iso-
topic ratio of the sample (e.g.: 18O/16O o 2H/1H) and that of a reference standard (RV-SMOW
where V-SMOW is the acronyms of Vienna Standard Mean Ocean Water) as shown in
Equation 1:
 
RSAMPLE
δð‰Þ ¼  1  1000 ð1Þ
RVSMOW

Where RSAMPLE = Isotopic content of the sample; RV-SMOW = Isotopic content of a stand-
ard sample.
During the monitoring time (June 2014 to July 2017) rain water showed an average δ18O
value of -8.5‰, while springs -10.2‰, wells -10.6‰ e piezometers -10.3‰. Minimum values of
-12.0‰ (rain water), -12.20‰ (springs), -12.3‰ (wells) e -12.1‰ (piezometers); maximum
values of -0.5‰ (rain water), -8.9‰ (springs), -8.45‰ (wells) e -8.7‰ (piezometers).
These data indicate a depletion of δ18O in groundwater compared to rain water. Further-
more, a low variability of the isotopic data emerges in groundwater (springs: 1σ = ±0.6‰;
wells: 1σ = ±0.9‰; piezometers: 1σ = ± 0.7‰) when compared to rain water (1σ = ±2.4‰).
The higher depletion that distinguishes groundwater, combined with lower variability of
isotopic data, suggests both a recharge occurring at higher altitudes and a homogenization
role played by the aquifers. This process tends to smooth the isotopic variability of ground-
water when compared to rain water precipitating in different periods of the year and at differ-
ent altitudes. In fact, at lower temperatures (winter) depleted values of δ18O are expected
while at higher temperature (summer) enriched values of δ18O are found.
Moreover, and due to the “altitude effect” (the progressive decrease in air temperature with
the elevation), an isotopic gradient can be found for rain water precipitated at different alti-
tudes. This makes possible to calculate the mean altitude at which rain water infiltrated and
thus estimate the groundwater flow paths length (Mazor, 1997). An average decrease of

538
-0.3‰/100 m for δ18O and of -2.5‰/100 m for δ2H is reported for the entire Italian territory
by Celico,1986. Other authors found different gradients (Longinelli, 2003; Penna, 2014) in
areas close to the Isarco valley.
Evaluating available isotopic data from rain collectors (2014–2017), it has been possible to
calculate the δ18O gradient for the area. A decrease of – 0.25‰/100m has been defined, in line
with those presented by the abovementioned Authors.
Data for monitored water points have been compared to the δ18O gradient line from rain
col- lectors as shown, for springs, in Figure 9.
Points falling closer to the line, as B_000234 located at 1600 m a.s.l., are supplied by small
recharge basins (between 1780 and 2040 m a.s.l.). On the contrary, points away from the line,
as B_000174 located at 880 m a.s.l., are supplied by larger hydrogeological basins where rain

Figure 9. Relation between isotopic values recorded at different springs (std. dev. ±1σ) and δ18O gradient
line from rain water.

Table 3. Estimated infiltration elevation for wells and piezometers.

Water table Estimated infiltration
ID Type elevation Average δ18O elevation Elevation difference

m. a.s.l. ‰ m. a.s.l. m. a.s.l.

B_000228 Well 555 -11.5 2060 1505
B_000376 Well 714 -11.0 1860 1146
B_000578 Well 662 -9.6 1300 638
B_000580 Well 774 -10.0 1460 686
B_000587 Well 483 -11.5 2060 1577
B_000605 Well 481 -11.2 1930 1449
B_000618 Well 680 -9.6 1300 620
B_000662 Well 522 -10.2 1540 1018
B_000837 Well 619 -10.0 1460 841
B_001244 Piezometer 941 -11.2 1940 999
B_001245 Piezometer 862 -10.3 1580 718
B_001252 Piezometer 758 -10.3 1580 822
B_001255 Piezometer 802 -11.2 1940 1138
B_001268 Piezometer 683 -9.7 1340 657
B_001270 Piezometer 681 -10.1 1500 819
B_001271 Piezometer 878 -10.6 1700 822
B_001353 Piezometer 614 -10.2 1540 926

539
water infiltrates at higher altitude respect to the spring point elevation (between 1600 and
2100 m a.s.l.). The same relation can be extended to wells and piezometers. In this case the
difference is rep- resented by a general higher difference in elevation (at least 600 m) between
the sampled points and their estimated recharge basins. This fact testifies, for piezometers, the
usually deeper and more developed groundwater flow paths. In case of wells, which are
always drilled through qua-ternary deposits along the river valleys, which are in turn fed from
higher elevation (Table 3).

6.2 UNNSTABLE ISOTOPES (3H)

Tritium is a hydrogen isotope that forms naturally in atmosphere. From 1960’s nuclear tests
significantly increased its concentration, so that it could have been used, due to its radioactive
nature (half-time T1/2 = 12.4 years), as a tracer in hydrogeology to estimate the time of water
from their infiltration to the discharge point.
It is possible to calculate the time of residence of water within the aquifer based on the
abundance of tritium with the following Equation 2:
 
T1=2

t¼  lnCp  lnCs ð2Þ
0:693

Where Cp = quantity of tritium in rain water (rain collectors) and Cs = quantity of tritium
in underground water.
Water with high concentration of Tritium infiltrated in recent times, while water with very
low quantities of this isotope can be considered older than 50–60 years, infiltrated before the
beginning of nuclear tests.
Variable concentrations of Tritium during different survey campaigns have been recorded.
Therefore, the residence time is calculated for every survey. In following table an overall aver-
age value is reported. This has been possible just for those water points which show a low
variability among different surveys. In fact, the maximum variation threshold for which data
can be considered as reliable, has been based on the highest instrumental error among the sur-
veys (2.5 TU, Tritium Units).

Table 4. Estimation of groundwater age.

Aver. ³H Aver. age Aver. ³H Aver. age
ID Type TU Years ID Type TU Years

B_001244 Piezometer <LD >50 B_000649 Spring 5 8

B_001245 Piezometer 7 3 B_000745 Spring 5 8
B_001248 Piezometer 6 5 B_000752 Spring 6 4
B_001252 Piezometer 6 5 B_000757 Spring 5 8
B_001255 Piezometer <LD >50 B_000761 Spring 6 5
B_001268 Piezometer 8 2 B_000828 Spring 5 11
B_001270 Piezometer 7 2 B_000834 Spring 5 8
B_001271 Piezometer <LD >50 B_000886 Spring 5 7
B_001353 Piezometer 7 3 B_001114 Spring 6 6
B_001372 Piezometer 2 23 B_001120 Spring 5 8
B_001382 Piezometer 5 8 B_001128 Spring 5 9
B_000227 Spring 5 9 B_001174 Spring 7 3
B_000235 Spring 5 8 B_000376 Well 6 6
B_000260 Spring 7 4 B_000578 Well -6 11
B_000263 Spring 6 6 B_000580 Well -4 10
B_000508 Spring 5 7 B_000587 Well 4 6
B_000585 Spring 5 9 B_000618 Well 13 4
B_000642 Spring 7 3 B_000662 Well 4 7

540
7 CONCLUSIONS

Most of spring water show low mineralization degrees which allow to consider them as origin-
ated by short groundwater flow paths: the latter are mainly developed within the altered and
fractured phyllite rock mass and the overlying quaternary fluvio-glacial deposits.
Only some areas are affected by more developed groundwater flows: the southern Laion-
Ponte Gardena area and the confluence between the Isarco and the Funes valleys.
The latter is characterized by the presence of a remarkable tectonic structure (Funes thrust)
that allows groundwater-mixing processes between shallow and deeper flow paths.
In Ponte Gardena area, water flows show higher mineralization due to the more extended
fracturing degree of the bedrock. Thanks to isotopic data, it has been possible to highlight the
presence of localized water flow-paths fed by higher altitudes. Nevertheless, low discharge
rates justify shallow groundwater flow paths.
Concerning piezometers, they usually show similar characteristics to those seen for springs.
Anyway, exceptions were found. Some of deeper water fall into the bicarbonate-alkaline
hydrofacies. Thanks to the isotopic analysis, these waters can be considered as the expression
of deeper groundwater flow paths developing within the rock mass.
However, it is important to highlight that no bicarbonate-alkaline water was found in
springs, wells and rivers which represent the final outlet of the groundwater from the area.
This is the focal point of the analysis due to their importance in terms of water resources
exploitation.
Therefore, for all the geological, hydrogeological and geochemical reasons explained above,
the forecasted impact of tunneling construction on water resources is expected to be very
limited.

REFERENCES

Celico, P. 1986. Prospezioni idrogeologiche Vol. 1. Napoli: Liguori Editore.

Civita, M. 2005. Idrogeologia applicata e ambientale. Milano: CEA.
Longinelli, A. & Selmo E. 2003. Isotopic composition of precipitation in Italy: a first overall map. Journal
of Hydrology 270: 75–88.
Mazor, E. 1997. Chemical and Isotopic Groundwater Hydrology. New York: Marcel Dekker Inc.
Penna, D., Engel, M., Dell’Agnese A. & Bertoldi, G. 2014. Tracer-based analysis of spatial and temporal
variations of water sources in a glacierized catchment. Hydrol. Earth Syst. Sci. 18: 5271–5288.

541