Ofterdinger 2004 Ground Water Flow Systems Rotondo Granit ETH 14108

I
J
Diss. ETH Nr. 14108
Ground water flow systems in the
Rotondo Granite, Central Alps
(Switzerland)
A dissertation submitted to the
SWISS FEDERAL INSTITUTE OF TECHNOLOGY (ETH)
ZRICH
for the degree of
DOCTOR OF NATURAL SCIENCE
presented by
ULRICH STEFAN OFTERDINGER
Diplom-Geologe
University of Kiel
born 18. November 1970
citizen of Germany
accepted on the recommendation of
Prof. Dr. Simon Lw, examiner
Prof. Dr. Wolfgang Kinzelbach, co-examiner
Prof. Dr. Klaus-Peter Seiler, co-examiner
Dr. Philippe Renard, co-examiner
PD Dr. Werner Balderer, co-examiner
Acknowledgements
I want to thank Prof. Simon Lw for admitting me in his group and having
guided this thesis. The progress of the project benefited very much from his
continuous encouragement and interest throughout this research. I am particular
grateful to Dr. Philippe Renard, co-examiner of this thesis, who always showed
great confidence in my work which was a constant source of motivation. His
inexhaustible willingness to discuss even basic problems helped me a great deal
in starting to understand ground water flow in fractured rocks in alpine regions.
I would also like to express my gratitude to the co-examiners PD. Dr. Werner
Balderer, Prof. Wolfgang Kinzelbach and Prof. Klaus-Peter Seiler for their help-
ful and encouraging remarks and advice on the interpretation of the project
results. I am particular grateful for the effort of all the examiners that allowed
me to meet my final deadline.
I am also thankful to the Furka-Oberalp-Bahn AG (Brig) and especially the
stationmasters at Realp for all their help and for providing me with the access
to the Bedretto-tunnel. Special thanks belong to Mr. Kurt Moser, who made all
the field work in the tunnel possible.
I want also to thank the responsibles at the meteorological stations who kindly
provided me with the precipitation sampIes.
I further wish to thank Mr. Roland Hallenbarter for letting me stay in the
Gerental-hut - an unforgettable experience of the field work in the Alps.
Special thanks go to my fellow PhD's in the Institute for the good times. In
11
particular I thank Volker Ltzenkirchen far his friendship and the 'discussions
across projects'. The common days in the field, in high altitudes and deep below
the ground will always be remembered.
I want to thank my family and Karina Fischer far their love and encourage-
ment. This work is dedicated to them.
Per aspera ad astra
Contents
Zusammenfassung
Summary
1 Introduction
xiii
xvi
1
2 Paper 1 - Environmental isotopes and hydrochemistry as indica-
tors for ground water flow systems in the Rotondo Granite 4
2.1 Introduction... 5
2.2 Geological setting 6
2.3 Hydrochemistry. 7
2.3.1 Spring discharges 7
2.3.2 Tunnel inflows. . 8
2.4 Environmental isotopes. 11
2.4.1 Tritium ..... 12
2.4.2 34Sj18Q in aqueous sulphate 14
2.4.3 2Hj18Q in precipitation, glacial meltwater and ground water 17
2.4.3.1 2Hj18Q in precipitation .. 17
2.4.3.2 2Hj18Q in glacial meltwater 18
2.4.3.3 2Hj18Q in ground water . 19
2.4.3.4 Lumped-parameter model 21
2.5 Discussion and Conclusion . . . . . . . . . 24
3 Paper 2 - Hydraulic subsurface measurements and hydrodynamic
modeling as indicators for ground water flow systems in the Ro-
tondo Granite 38
3.1 Introduction.............................. 39
III
IV
3.2 Geological and hydrogeological setting ..... 40
3.2.1 Major geological and hydrogeological units 40
3.2.2 Structural geology 42
3.2.3 Isotopes/ Chemistry . 43
3.2.4 Spring line. . . . . . 44
3.3 Hydrogeological parameters 44
3.3.1 Hydraulic heads . . . 44
3.3.2 Ground water fluxes 45
3.3.3 Hydraulic conductivity 46
3.3.4 Ground water recharge rates 49
3.3.5 Flow porosities 51
3.4 Conceptual model . . . 52
3.5 Numerical Model ... 53
3.5.1 Modeling approach 54
3.5.2 Model calibration . 56
3.5.2.1 Scenario 1 56
3.5.2.2 Scenario 2 . 56
3.5.2.3 Scenario 3 . 57
3.5.3 Model results .. 57
3.5.4 Sensitivity study 59
3.5.5 Discussion of model results . 61
3.6 Comparison to results from environmental isotope study 66
3.7 Conclusion . . . . ................... . . 70
4 Conc1uding remarks 101
References 105
Appendices
A Structural geology
B Investigations related to the research borehole
B.l Borehole location and construction
B.2 Geophysical borehole logging ....
121
121
125
126
127
v
B.3 Hydraulic Packer-testing . . . . . . . . . . . . . . . . . . . . . . . 135
B.4 Borehole completion and long-term monitoring of hydraulic pres-
sures in borehole intervals . 140
B.4.1 Borehole completion 140
B.4.2 Long-term hydraulic pressure record 142
B.4.3 Effects of tidal forcing . . . . . . . . 144
C Hydrochemistry
D Environmental isotopes
E Discharge measurements
149
164
173
F Dimensionless solution for constant head test in ideal aquifer
with a well of finite radius and constant head boundary 176
Curriculum Vitae 178
List of Figures
2.1 Enlarged plan view of the research area . . . . . . . . . . . . . .. 29
2.2 Record of automated discharge measurements at the stations 81
and 82 in the Bedretto-tunnel . . . . . . . . . . . . . . . . . . .. 30
2.3 Record of annual precipitation at the meteorological station Grim-
seI/Hospiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 30
2.4 Mean temperature, tritium concentration and 15
18
0 content of col-
lected ground water sampIes along the Bedretto-tunnel . . . . .. 31
2.5 Plot of tritium versus sodium content for ground water sampIes
from the Bedretto-tunnel. 32
2.6 15
34
8 values of some geologically important materials in comparison
to the encountered range of 15
34
8 in the Rotondo Granite. . . . .. 32
2.7 Commonly observed ranges of 15
34
8 and 15
18
0 values for sulphates
in relation to measurements from the Rotondo Granite. . . . . .. 33
2.8 Effects of various processes on the 15
18
0 values of 8 0 ~ - and H
2
0. 33
2.9
18
0H
2
0 versus 15180804 values from published data in comparison
to data from the Rotondo Granite. . . . . . . . . . . . . . . . .. 34
2.10 Detailed plot of 15180804 versus 15348804 in aqueous sulphate in
sampIes from the Bedretto-tunnel . . . . . . . . . . . . . . . . .. 35
2.11 Local meteoric water line based on
2
H/
18
0 data in precipitation
at the meteorological stations together with the meteoric water
line and
2
H/
18
0 data from ground water sampIes collected in
the Bedretto-tunnel . . . . . . . . . . . . . . . . . . . . . . . . .. 35
2.12 Relationship between altitude and mean annual 15
18
0 content in
precipitation at meteorological stations in the vicinity of the re-
search area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 36
VI
VB
2.13 Observed 8
18
0 data at the station T9 and calculated mean signal
of remaining ground water sampling stations together with the
results from the lumped-parameter model. . . . . . . . . . . . .. 37
3.1 Schematic plan view of the model domain indicating the projected
trace of the subsurface galleries and the individual geological units
within the domain 79
3.2 Enlarged plan view of the model domain . . . . . . . . . . . . .. 80
3.3 Long-term record of hydraulic pressures in research borehole . .. 81
3.4 Schematic trace ofthe Furka-basetunnel and Bedretto-tunnel with
indicatars for the direction of tunnel drainage and tunnel metrics 81
3.5 Map view of the distributed ground water recharge rates applied
in the ground water flow model. . . . . . . . . . . . . . . . . . .. 82
3.6 Enlarged map view of the distributed ground water recharge rates
applied in the ground water flow model . . . . . . . . . . . . . .. 83
3.7 Schematic overview of the layer structure applied in the 3d finite
element model . . . . . . . . . 84
3.8 3d view of finite element mesh 85
3.9 Plot of measured cumulative ground water inflow to the Bedretto-
tunnel along the tunnel profile versus simulation results far the
individual model scenarios . . . . . . . . . . . . . . . . . . . . .. 86
3.10 Depth to the ground water table below ground surface for model
scenario 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87
scenario 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 88
3.12 Enlarged plan view of the upper Gerental-valley with the position
ofthe recharge areas to the Bedretto-tunnel for the model scenarios
and model variant 1&2.. . . . . . . . . . . . . . . . . . . . . . .. 89
3.13 Range of advective travel times to the Bedretto-tunnel for 120m
intervals along the tunnel axis (nfRG = 2.0 x 10-
4
) ..... 90
3.14 Range of advective travel times to the Bedretto-tunnel for 120m
intervals along the tunnel axis (nfRG = 1.0 x 10-
3
) 90
3.15 Plan view of the model domain with isolines of drawdown due to
the tunnel galleries far model scenario 1 91
Vlll
3.16 2d section of hydraulic potentials for model scenario 3 as a response
to distributed recharge rates and the presence of the subsurface
gaIlery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 92
variant 1 with uniform ground water recharge of 5.75x10-
4
m/d. 93
4
m/d. 94
3.19 2d section of hydraulic potentials for model scenario 3 indicating
local and regional ground water flow systems . . . . . . . . . . .. 95
3.20 2d section of hydraulic potentials for model scenario 2 and 1 in-
dicating the influence of the large scale fault zone on the head
distribution in the Saashorn area. . . . . . . . . . . . . . . . . .. 96
3.21 2d section of hydraulic potentials for model scenario 2 and 1 in-
dicating the influence of the large scale fault zone on the head
distribution in the Leckihorn area. . . . . . . . . . . . . . . . . .. 97
3.22 Typecurves for constant head test in finite radius weIl with con-
stant head boundary foIlowing Murdoch and Franco (1994) . . .. 98
3.23 Resdidual percentage between qd and ql versus t
d
for constant head
test in a finite radius weIl with a constant head boundary at dis-
tance h
d
= 800 99
3.24 Cross-section along the profile of the Bedretto-tunnel with loca-
tion of sampled inflows and mean measured sampIe temperature,
tritium content and 6"
18
0 value 100
A.1 Stereographic projection of the scanline data . . . . . . . . . . . . 122
A.2 Azimuth-frequency histograms of fracture orientations measured
in the Bedretto-tunnel . . . . . . . . . . . . . 123
A.3 Organisation of data storage for scanline data 124
B.1
B.2
B.3
BA
B.5
B.6
Location of research borehole . . . . . . . . .
Diagnostic plot of Th/K-ratio for the 'Y-spectroscopy data
Geophysical borehole logs . . . . . . . . . . . . . .
Fluid-logging results .
Packer test results for borehole interval 99.2-93.0 m
Packer test results for borehole interval 9 9 . 2 ~ 6 3 . 5 m
126
131
133
134
136
136
IX
B.7 Packer test results for borehole interval 99.2-65.6 m . . . . . . . . 137
B.8 Packer test results for borehole interval 99.2-65.6 m (log-log plot
with match of GRF-model) 138
B.9 Packer test results for borehole interval 99.2-65.6 m (semi-Iog plot
with match of GRF-model) 139
B.10 Organisation of data storage - Packer-tests 140
B.11 Borehole completion for long-term monitoring 141
B.12 Long-term record of hydraulic pressures in research borehole 142
B.13 Long-term average ground water recharge pattern in the upper
Gerental-valley . . . . . . . . . . . . . . . . . . . . . . . . . 143
B.14 Organisation of data storage - Long-term monitoring record 144
B.15 Comparison of barometrie and hydraulic pressures . . . . . . 144
B.16 Power spectrum density plot for hydraulic pressures in intervall 145
B.17 Power spectrum density plot for hydraulic pressures in interval 2 146
B.18 Residual pressure variations in interval 1 in the diurnal frequency
range. 147
B.19 Residual pressure variations in interval 1 in the semi-diurnal fre--
quency range. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
D.1 Organisation of data storage - environmental isotopes in precipi-
tation at meteorological stations Guttannen, Meiringen and Grim-
sel/Hospiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
E.1 Organisation of data storage - continuous discharge measurements
at the stations 81 and 82 in the Bedretto-tunnel . . . . . . . . . . 175
List of Tables
2.1 Mean orientat ion of fracture sets and corresponding mean normal
fracture frequency . . . . . . . . . . . . . . . . . . 27
2.2 Mean in situ parameters of sampled ground water 28
3.1 Mean orientation of fracture sets and corresponding mean normal
fracture frequency . . . . . . . . . . . . . . . . . . . . . . . . . .. 73
3.2 Integral inflow rate measurements in the Bedretto- and Furka-
basetunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., 74
3.3 Analytical approximations of hydraulic conductivities in the Ro-
tondo granite . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 74
3.4 Initial approximations of effective hydraulic conductivities for the
geological units within the model domain together with calibrated
hydraulic conductivities for the individual scenarios . . . . . . .. 75
3.5 Calibration targets for inflow rates along specific tunnel sections
and hydraulic heads in monitoring intervals with calibrated values
for the individual model scenarios . . . . . . . . . . . . . . . . .. 76
3.6 Approximated advective travel times across recharge areas for the
model scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77
3.7 Simulated ground water inflows to the tunnel galleries and sim-
ulated heads in the monitoring intervals for model variants with
uniform ground water recharge rates 78
B.1 Transmissivities and hydraulic conductivities from packer tests in
the research borehole . . . . . . . . . . . . . . . . . . . . . . . . . 139
C.1 Location of ground water sampling sites at the terrain surface for
hydrochemical analysis . . . . . . . . . . . . . . . . . . . . . . . . 151
x
Xl
Co2 Location of ground water sampling sites in the Bedretto-tunnel for
hydrochemical analysis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 151
Co3 In situ parameters and chemical analysis of ground water sampIes
collected from springs in the Rotondo area 0 0 0 0 0 0 0 0 0 0 0 0 152
CA In situ parameters and chemical analysis of ground water sampIes
collected in the Bedretto-tunnel at sampling location Tl 0 0 0 0 0 153
collected in the Bedretto-tunnel at sampling location T2 0 0 0 0 0 154
C.6 In situ parameters and chemical analysis of ground water sampIes
collected in the Bedretto-tunnel at sampling location T5 . 0 0 0 157
collected in the Bedretto-tunnel at sampling location T6 0 0 0 0 158
collected in the Bedretto-tunnel at sampling location T7 0 0 0 0 159
Coll In situ parameters and chemical analysis of ground water sampIes
collected in the Bedretto-tunnel at sampling location T9 0 0 0 160
collected in the Bedretto-tunnel at sampling location Tll 0 0 0 0 161
collected in the Bedretto-tunnel at sampling location SI 0 0 0 0 0 162
collected in the Bedretto-tunnel at sampling location SI 0 0 0 0 163
Do1 Location ofthe meteorological stations Guttannen, Meiringen, Grim-
seI/Hospiz, Binn, Oberwald, Andermatt, Gtsch and Robiei 0 0 166
Do2 Tritium concentration in precipitation at meteorological stations
Guttannen, Meiringen and Grimsel/Hospiz (1990's) 0 0 0 0 0 0 167
D.3 Tritium concentration in ground water sampIes from the Bedretto-
tunnel .. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 168
DA 8
34
S and 15
18
0 values in aqueous S O ~ - in ground water sampIes
collected in the Bedretto-tunnel. .. 0 0 0 0 0 0 0 0 0 0 0 0 0 168
Xll
D.5 b
18
0 and b
2
H values in precipitation at the meteorological stations
Binn, Oberwald, Andermatt, Gtsch and Robiei . . . . . . . . . . 169
D.6 b
18
0 and b
2
H values in precipitation at meteorological stations
Guttannen, Meiringen and GrimseljHospiz (1970-1999) . . . . . . 170
D.7 b
18
0 and b
2
H values in ground water sampIes collected in the
Bedretto-tunnel . . . . . . . . . . . . . . . . . . . . . . . . . .. 171
D.8 b
18
0 and b
2
H values in ice sampIes collected on the Gerenglacier 172
E.l Integral flow rate measurements in the Bedretto-tunnel .
E.2 Integral flow rate measurements in the Furka-basetunnel
174
175
Zusammenfassung
Die Untersuchung regionaler Grundwasserfliessysteme in geklfteten Kristallin-
gesteinen alpiner Regionen stellt aufgrund der komplexen Wechselwirkungen zahl-
reicher Einflussgrssen eine grosse Herausforderung dar. Zu diesen Einflussgrssen
zhlen unter anderem die Geologie und Topographie, die Grundwasserneubil-
dungsbedingungen, Untertagebauwerke und strukturgeologische Aspekte wie
Trennflchengefge und regionale Strzonen. Die Motivation dieser Disserta-
tion liegt zum einen in der Charakterisierung der Grundwsser im geklfteten
Kristallin hochalpiner Regionen im Hinblick auf Ihre Zusammensetzung, Ur-
sprung und Neubildungsbedingungen. Hierfr wird die hydrochemische und iso-
topische Zusammensetzung der Grundwsser untersucht. Eine weitere Fragestel-
lung ist die Bedeutung einzelner Einflussgrssen auf die Grundwasserfliessysteme
im Untersuchungsgebiet. Diese werden im Rahmen eines regionalen numerischen
Grundwasserstrmungsmodells untersucht. Die Ergebnisse der Modellsimulatio-
nen werden anschliessend im Hinblick auf die Ergebnisse der Isotopen- und Hy-
drochemiestudie evaluiert. Diese Untersuchungen haben zur Erstellung von zwei
Publikationen gefhrt, die im folgenden zusammmengefasst sind.
Zur Untersuchung der Grundwasserneubildungsbedingungen und Fliessyste-
me in geklfteten Graniten eines hochalpinen Untersuchungsgebiets in den
schweizer Zentralalpen wurden Messungen zur Hydrochemie und zum Gehalt
an Umweltisotopen (6
18
0, 6
2
H, Tritium sowie 6
34
S/6
18
0 in gelstem S O ~ - ) in
Grundwasser- und Niederschlagsproben durchgefhrt. Die Messungen in Grund-
wasserproben aus einem tiefgelegenen Tunnel weisen auf die Bedeutung von akku-
mulierten Winterniederschlgen und glazialen Schmelzwssern bei der Grund-
wasserneubildung hin. Die hydrochemische und isotopische Zusammensetzung
der Tunnelwsser zeigt nur geringe zeitliche Schwankungen. Die 6
18
0_Werte im
Xlll
XIV
Niederschlag zeigen saisonale Variationen von rv14 %0 und eine Hhenabnahme
von 0.23 (0.03) %0 pro 100 Hhenmeter. 6"
2
H- und 6"
18
0-Werte im Niederschlag
sind deutlich korreliert und liegen nah an der Niederschlagsgeraden (Meteoric
Water Line). 6"
18
0_ und 6"
2
H-Werte in Grundwasserproben verweisen durch eine
vergleichbare Korrelationen auf den meteorischen Ursprung dieser Wsser. Die
negativeren 6"
18
0_Werte im Grundwasser im Vergleich zu den 6"
18
0_Werten des
Niederschlags im Zeitraum der Grundwasserneubildung deutet auf einen sig-
nifikanten Beitrag von akkumulierten 180-abgereicherten Winterniederschlgen
bei der Grundwasserneubildung hin. Die hydrochemische Zusammensetzung des
Grundwassers (Na-Ca-HC0
3
-S0
4
(-F)) spiegelt die chemische Entwicklung des
Grundwassers entlang des Fliesspfades durch den Granitkrper wieder. Die Aus-
richtung der 6"
34
S/6"
18
0 Messwerte entlang einer Mischungsgeraden mit der Stei-
gung rvO.6 weist lokal auf einen potentiellen Beitrag von Grundwssern aus den
benachbarten Gneisseinheiten hin. Gemessenen Tritiumwerte im Grundwasser
liegen zwischen 2.6 und 16.6 TU, wobei die kleinsten Werte mit einer lokalen
negativen Temperaturanomalie und abgereicherten
18
0_Werten assoziiert sind.
Diese stehen mit dem lokalen Beitrag submoderner glazialer Schmelzwsser an der
Grundwasserneubildung im Zusammenhang. Diese lokale Grundwasserneubil-
dung erfolgt entlang bevorzugter Fliesspfade innerhalb des Granitkrpers, welche
in der Regel von hydraulisch aktiven Scherzonen und kataklastischen Strzonen
gebildet werden.
Regionale Grundwasserfliessysteme in alpinen Gebieten werden durch eine
Vielzahl von Faktoren beeinflusst. Prgende Faktoren sind dabei unter anderem
die Geologie und Topographie, die Grundwasserneubildungsbedingungen, struk-
turgeologische Aspekte wie Trennflchengefge und regionale Strzonen sowie
Untertagebauwerke. Mit Hilfe eines regionalen numerischen Grundwasserstr-
mungsmodells untersuchen wir den Einfluss eines tiefgelegenen Tunnels und einer
Modell-Strzone auf das Grundwasserregime im geklfteten Rotondo-Granit. Des
weiteren untersuchen wir den Einfluss rumlich verteilter Grundwasserneubil-
dungsraten auf das Fliessystem und ins besondere auf die Lage des freien Grund-
wasserspiegels. Die Modellergebnisse zeigen deutliche ungesttigte Bereiche mit
einer Mchtigkeit von bis zu mehreren hundert Metern unterhalb der Gebirgszge
im Untersuchungsgebiets. Die Tunnelbauwerke haben eine deutliche Auswirkung
xv
auf die Druckhhenverteilung im Modellgebiet und fhren dabei lokal zu einer
Umkehr der natrlichen Druckhhengradienten. Der Einfluss rumlich verteilter
Grundwasserneubildungraten wird in Hinblick auf die Lage der Einzugsgebiete
zum Tunnel und den damit verbundenen Ursprung der Grundwasserneubildung
der Tunnelwsser (rezenter Niederschlag oder Gletscherschmelze) und im Hin-
blick auf die Lage des Grundwasserspiegels untersucht. Zum Vergleich werden
Modellrechnungen mit verschiedenen rumlich einheitlichen Neubildungsraten
durchgefhrt. Die Lage des Grundwasserspiegels innerhalb exponierter, stark re-
liefierter Gebirgszge zeigt dabei ein sensitiveres Verhalten auf Vernderungen der
Grundwasserneubildungsraten und Permeabilitten als innerhalb von Gebirgsz-
gen mit geringerem topographischem Relief. Die Modell-Strzone hat im konzep-
tionellen Rahmen der numerischen Simulationen nur einen geringen Einfluss auf
die Lage des Grundwasserspiegels. Sie fungiert primr als rascher Fliesspfad fr
Grundwasser aus den Neubildungsgebieten der Gletscherregion zum Tunnel. Dies
steht in bereinstimmung mit der Analyse der Umweltisotope, die ebenfalls einen
Beitrag glazialer Schmelzwsser zur Grundwasserneubildung anzeigt.
Die erste Publikation zeigt auf, dass die Analysen der Umweltisotope In
den Stollenwssern wertvolle Hinweise auf die Grundwasserneubildungsbedin-
gungen der tiefen Grundwassersysteme im Untersuchungsgebiet liefern. In der
zweiten Publikation wird gezeigt, dass bereits numerische Simulationen mit einem
vereinfachten konzeptionellen Ansatz grundlegende Einsichten in die Wechsel-
wirkung einzelner Einflussgrssen auf die regionalen Strmungssysteme in der
untersuchten Hochgebirgsregion erlauben und dabei die auf den hydrochemischen
und isotopischen Analysen beruhenden Interpretationen untersttzen.
Summary
The assessment of regional flow systems in fractured crystalline rocks in high
alpine regions poses a great chaHenge due to the complex interactions of a number
of controlling factors such as structural geology, topography, recharge conditions
and man-made underground structures. The motivation of this thesis is first to
characterize the ground water in fractured crystalline rocks in a high alpine re-
gion by means of hydrochemistry and environmental isotope analysis as weH as
hydraulic measurements. Secondly, the impact of a number of controlling factors
on the ground water flow systems are investigated in a regional numerical ground
water flow model. The results of these simulations are then also evaluated with
regard to the findings from the isotope and hydrochemistry analysis. These stud-
ies have led to two papers, whose abstracts are presented below.
To investigate ground water recharge conditions and circulation systems in
fractured granites of a high alpine catchment in the Central Alps (Switzerland),
measurements of environmental isotope ratios, hydrochemical composition and
in situ parameters of ground water were performed in a deep tunnel. These
measurements demonstrate the significance of accumulated winter precipitation
and glacial meltwater to the recharge of deep ground water flow systems. Hydro-
chemical and in situ parameters as weH as
18
0 in ground water sampIes coHected
in the tunnel show only smaH temporal variations. The precipitation record of
18
0 shows seasonal variations of rv14 %0 and a decrease of 0.23 (0.03)%0 per
100 m elevation gain.
2
H and
18
0 in precipitation are weH correlated and plot
elose to the Meteoric Water Line and likewise
2
H and
18
0 in ground water
sampIes, reflecting the meteoric origin of the latter. The depletion of
18
0 in
ground water compared to
18
0 content in precipitation during the ground wa-
ter recharge period indicates significant contributions from accumulated depleted
XVI
XVll
winter precipitation to ground water recharge. The hydrochemical composition of
the encountered ground water (Na-Ca-HC0
3
-S0
4
(-F)) reflects an evolution ofthe
ground water along the flow path through the granite body while the alignment
of the 5
34
S/5
18
0 data along a mixing line with a slope of 0.6 indicates locally
a potential minor contribution of a ground water component originating in the
neighboring gneiss formation. Observed tritium concentrations in ground water
range from 2.6 to 16.6 TU with the lowest values associated with a local negative
temperature anomaly and anomalous depleted
18
0 in ground water, demonstrat-
ing the effect of local ground water recharge from meltwater of submodern glacial
ice. Such localized recharge from glaciated areas occurs along preferential flow
paths within the granite body which are mainly controlled by observed hydraulic
active shear fractures and catac1astic faults.
Regional ground water flow in high mountainous terrain is governed by a
multitude of factors such as geology, topography, recharge conditions, structural
elements such as fracturation and regional fault zones as weIl as man-made under-
ground structures. By means of a numerical ground water flow model we consider
the impact of deep underground tunnels and of an idealized major fault zone on
the ground water flow systems within the fractured Rotondo granite. The posi-
tion of the free ground water table as response to the above subsurface structures
and in particular with regard to the influence of spatial distributed ground water
recharge rates is addressed. The model results show significant unsaturated zones
below the mountain ridges in the study area with a thickness of up to several
hundred meters. The subsurface galleries are shown to have a strong effect on
the head distribution in the model domain, causing locally areversal of natu-
ral head gradients. With respect to the position of the catchment areas to the
tunnel and the corresponding type of recharge source for the tunnel inflows (i.e.
glaciers or recent precipitation) as weIl as water table elevation the influence of
spatial distributed recharge rates is compared to uniform recharge rates. Water
table elevations below the weIl exposed high-relief mountain ridges are observed
to be more sensitive to changes in ground water recharge rates and permeability
than below ridges with less topographic relief. In the conceptual framework of
the numerical simulations the model fault zone has less influence on the ground
water table position but more importantly acts as fast flow path for recharge
XVlll
from glaciated areas towards the subsurface galleries. This is in agreement with
the previous study, where the imprint of glacial recharge was observed in the
environmental isotope composition of ground water sampled in the gallery.
The first paper shows that in the research area the analysis of the environ-
mental isotope content of ground water provides valuable information on the
origin of the ground water and the recharge conditions of deep ground water flow
systems. The second paper demonstrates that numerical simulations based on a
simplified conceptual approach already yield insight into the influence of different
controlling fact ars on the regional flow systems and provide furt her confidence
in the interpretations based on the analysis of the hydrochemical and isotopic
composition of the ground water.
1
Introduction
The assessment of regional ground water flow in fractured crystalline rocks in
high mountainous regions poses a great challenge due to the complex interaction
of controlling factors. T6th (1963) and Forster and Smith (1988a) have shown
that topography, i.e. relief, three-dimensional shape and basin scale strongly gov-
ern the pattern of regional ground water flow systems in mountainous terrain.
In fractured and faulted rocks additional complexity is introduced by hydraulic
anisotropy due to preferred orientations of water conducting fractures and by
the presence of regional fault zones providing preferential large scale flow paths
(L6pez and Smith 1995). Recharge conditions across catchments are strongly
variable making the assessment of ground water flow patterns and water table
elevations even more difficult. Furthermore, man-made structures such as sub-
surface galleries may have major impacts on the ground water flow regime by
modifying the natural flow field (Monjoie 1990; Loew et al. 1996).
This dissertation is motivated by the prospect of increasing our understand-
ing of regional ground water flow systems in fractured rocks in alpine regions by
combining the analysis of surface hydrological processes, structural geology, hy-
drochemistry and environmental isotope composition of the ground water as well
as hydraulic measurements. For this investigation, a research site in the Rotondo
area of the Central Swiss Alps was chosen that allowed a multitude of observa-
tions and measurements both on the terrain surface and in the subsurface along
an unlined tunnel. On the basis of the collected data we aim to characterize the
regional ground water flow regime in terms of a consistent interpretation of the
1
2
different data sets. The acquired data is furthermore used to constrain a regional
numerical ground water flow model, in which the significance of individual con-
trolling factors on the flow system is evaluated. In particular we are interested to
assess the impact of spatial varying recharge conditions and recharge sources on
the flow system. Furthermore, we investigate the influence of a simplified large
scale fault zone and of the subsurface gallery. The position of the free water table
in response to the above influencing factors is another focus of theses evaluations.
The thesis consists of two papers, which are self-contained and can be read
independently, and concluding remarks. Details of the collected field data and
field investigations are given in the appendices and on the attached CD-ROM.
In the first paper we investigate the hydrochemical composition and environ-
mental isotope content (34Sj1
S
0 in aqueous S O ~ - , tritium and 2Hj1
s
0) of the
encountered ground water in the research area. Fontes et al. (1986) and Dubois
(1993) have shown that hydrochemical and isotopic analysis of ground water in
crystalline rocks provide a feasible tool to delineate the origin and recharge con-
ditions of theses waters. Specific recharge sources such as snowpack and glaciers
(Martinec et al. 1982) in high alpine catchments yield characteristic signatures
in terms of environmental isotope content (Abbott et al. 2000; Flerchinger et al.
1992; Ward et al. 1999), which are reflected in the composition of the ground
water encountered along the subsurface gallery. In the presented paper, analysis
and interpretation are focused on identifying the origin and mean residence times
of the encountered ground water, on assessing the impact of the large scale fault
zone on ground water flow to the subsurface gallery and on extracting informa-
tion on the ground water recharge conditions in the study area.
In the second paper we use the collected hydraulic and structural field data
to constrain a regional ground water flow model in an equivalent continuum
approach. Spatial distributed recharge rates, extracted from the results of a
hydrological model of the research area (Vitvar and Gurtz 1999), are applied as
upper boundary condition of the model and model results are discussed in terms
of consistence with the findings from the first paper. Furthermore the impact of
the large scale fault zone, the subsurface gallery and varying recharge conditions
3
are addressed with regard to the position of the ground water table and recharge
areas to the gallery.
2
Environmental isotopes and
hydrochemistry as indicators for
ground water flow systems in the
Rotondo Granite
u. s. Ofterdinger, w. Balderer, S. Loew & Ph. Renard
Submitted to
Ground Water
4
5
2.1 Introduction
High alpine catchments in heterogeneous fractured crystalline rocks present a
great challenge when assessing the large scale ground water flow systems within
these formations. Regional flow systems in mountainous terrain have been the-
oretically addressed by e.g. T6th (1963, 1984), Forster and Smith (1988a,b)
and Zijl (1999) and the impact of fault zones on such systems were studied by
e.g. L6pez and Smith (1995). Amongst others, studies in crystalline massifs of
Switzerland and France show the influence of the degree of fracturation and their
orientation on preferential flow paths on the massif scale (Lhomme et al. 1996;
Pistre 1993). Crucial factors for understanding the nature of the flow systems are
the analysis of the origin of encountered ground water and especially the recharge
conditions in these topographically complex terrains. Systematic studies of the
chemical composition of ground water encountered in crystalline formations of
the Swiss Alps show how these parameters can assist in delineating the origin of
the encountered waters (Dubois 1993). Recharge conditions in these high moun-
tainous areas are strongly influenced by seasonal variations (e.g. Kattelmann and
EIder (1991)). Accumulated winter precipitation in the snowpack and meltwater
from glaciated areas both contribute to ground water recharge (Martinec et al.
1982; Forster and Smith 1988a). The influence of the surface hydrology and in
particular of snowmelt and glaciation on shallow ground water systems has been
investigated and a characteristic signature of this recharge source was shown in
terms of environmental isotope content and hydrodynamic response (Abbott et al.
2000; Flerchinger et al. 1992; Ward et al. 1999). However, these investigations
are mainly restricted to shallow ground water systems in the decompressed zone
of the basement rock (Cruchet 1985) with near surface weathering and stress
relief.
To investigate the effects of the specific recharge conditions in high alpine
regions on the deeper flow systems within fractured crystalline rocks, a high
relief catchment (total area rv40 km
2
) in the western Gotthard-massif was chosen
for this study. Results from a hydrological model (PREVAV-ETH) applied to
the research area provide estimates of spatially and temporal distributed ground
water recharge rates (Vitvar and Gurtz 1999), indicating high recharge rates in
the glaciated areas during ablation periods. Even though only a small percentage
(rv 14.6%) of the total catchment is glaciated, significant effects on the ground
6
water flow regime are anticipated, especially regarding the flow to a deep and
unlined tunnel (Bedretto-tunnel), which partly passes beneath the research area
and its glaciated regions. This tunnel was used for direct subsurface observations
and monitoring of ground water flow and sampling. Analysis of the hydrochemical
composition as weIl as the environmental isotope content was used to establish
the relationship between surface waters, spring discharges and deeper ground
water. Interpretation of the data was furthermore based on the analysis of the
structural anisotropy and heterogeneity of the host rock due to fracturation and
faulting.
2.2 Geological setting
The research area is situated in the western Gotthard-Massif of the Swiss Alps.
Geologically it consists mainly of the late Hercynian Rotondo-Granite (
rv
220 Mio
years; Jger and Niggli (1964)) and orthogneiss of early Palaeozoic age (rv420
Mio years; Arnold (1970)). Topographie elevation of the granite body ranges
from 1800 to 3200 m a.s.l. (Figure 2.1).
On the northern margin ofthe study area the Furka-basetunnel passes through
the granite body at an elevation of rv1490 m a.s.l. An abandoned and unlined
support segment to this basetunnel, the Bedretto-tunnel, is furthermore passing
through the granite body in northwest-southeast direction. Due to the partial
collapse of the tunnel, only the northern rv1.5 km are still accessible. Figure
2.1 shows in particular that this latter tunnel segment is also passing beneath
glaciated areas of the study domain. It offers the opportunity for direct subsur-
face observations and was used for sampling of ground water and monitoring of
discharges and in situ parameters. The granite body shows varying degrees of
fracturation and faulting throughout the research domain, changing both in lat-
eral and vertical directions. Data on fracture orientation and fracture frequency
were gathered by means of scanline surveys (Priest 1993) on surface outcrops
and along the unlined tunnel profile of the Bedretto-tunnel. From this data four
major fracture sets can be deduced (Table 2.1), where especially set 1 and 2 are
dominant at the terrain surface. The NW-SE striking fracture sets (3 and 4) are
more abundant along the tunnel profile and are believed to be influenced by the
tunnel construction, i.e. induced or reactivated by local stress amplifications.
7
In addition to the steeply dipping sets described above, further shallow dip-
ping fractures can be identified at the terrain surface. These latter fractures
generally strike sub-parallel to the valleys with dip directions towards the valley
bottom, most probable representing quaternary stress release joints. Occasion-
ally, concentrical jointing can also be observed within the granite at the terrain
surface. In addition to the fracturation, steeply dipping fault zones can be iden-
tified as important larger features, both at surface and along the tunnel profile.
They are mostly orientated sub-parallel or at acute angles to the general trend
of the alpine structures of the Gotthard-Massif (ENE-WSW). One of the ma-
jor fault zones intersecting the Bedretto-tunnel is indicated on Figure 2.1. Both
preferred orientations of fracturation (strike ENE-WSW) and on a larger scale
the orientation of fault zones cause a significant structural anisotropy within the
granite body.
2.3 Hydrochemistry
SampIes of the encountered ground water were collected from spring discharges
at the terrain surface of the granite body as weIl as from inflows to the Bedretto-
tunnel. Their major ionic composition was analyzed by ion-exchange liquid chro-
matography while parameters such as pR, temperature, conductivity, alkalinity
and oxygen content were measured in situ during sampling with electronic probes
or field titration respectively. Saturation indices (SI=log IAP/K
T
) for several
minerals where calculated with WATEQ4F (Ball and Nordstrom 1992).
2.3.1 Spring discharges
Few spring discharges from the granite body can be identified in the research
area. They are mostly associated with zones of denser fracturation with NE-SW
striking, steeply dipping fractures (Figure 2.1). Table 2.2 gives an overview of
the average in situ parameters measured at these spring discharges during three
successive summer field campaigns (1997-1999). In terms of decreasing meq%
of cations and anions respectively, the encountered waters are characterized as
dilute, slightly alkaline Ca-Na-RC0
3
(-S04)-waters. Their mean Ca/Na ratio,
deduced from meq/l-concentrations, lies at 2.58 (0"=0.9). All encountered wa-
ters are undersaturated with regard to calcium bearing minerals such as Calcite
8
or sulphate-minerals such as Gypsum or Anhydrite (mean value for SICalcite=
-3.417 (T=0.122; SIGypsum=-4.793 (T=0.151; SIAnhydrite=-4.541 (T=0.151). Spring
discharge from the northern orthogneiss shows a different hydrochemical composi-
tion and can be characterized as Ca-HC0
3
-(S04)-waters with a mean Ca/Na ratio
of 8.93 ((T=1.26). Mean in situ parameters are stated in Table 2.2. Compared to
the spring waters encountered from the granite body the discharge from the or-
thogneiss is less but still undersaturated with regard to calcium bearing minerals
or sulphate-minerals (mean value for SICalcite=-1.756 (T=0.276; SIGypsum=-3.691
(T=0.046; SIAnhYdrite=-3.948 (T=0.045).
2.3.2 Tunnel inflows
Along the rv1.5 km accessible section of the unlined Bedretto-tunnel, eleven
tunnel inflows were monitored and sampled (Figure 2.1). Hydrochemical sam-
pling was carried out from September 1998 until November 1999 (15 months)
on a monthly or bi-monthly basis, while in situ parameters were measured from
September 1998 until August 2000. Measurements of discharge rates at the in-
dividual sampling locations were carried out manually with calibrated vessels.
For measurements of integral inflow rates along particular tunnel sections, depth-
integrated flow velocity measurements with an anemometer were carried out along
cross-sections in the basal drain of the tunnel, which cumulatively captures the
tunnel inflows. Discharge was then calculated according to IS0748:1997(E) (ISO
1997). At the specific observation points SI and S2 (Figure 2.1), triangular-
notch thin-plate weirs, according to ISOI438/1:1980(E) (ISO 1980), equipped
with float level gauges, were installed for continuous measurement of the co1-
lected discharge at these locations. Measurements were recorded from January
1999 until August 2000. However, the records prior to mid-June 1999 were partly
obscured by technical problems and damages caused by rock-bursts in the gallery.
Discharge rates were calculated from the records applying the Kindsvater-Shen
formula (ISO 1980).
Similar to the surface observations, inflows are mainly associated with NE-SW
striking fractures and/or fault zones. This indicates that the observed dominant
fracture orientations not only cause a strong structural anisotropy within the
granite body but are furthermore causing also a hydraulic anisotropy and hetero-
geneity favoring preferential flow in the plane of these features. The observation
9
points T9 and SI are associated with a major steeply dipping fault zone, striking
rv75 (Figure 2.1). This approximately 120 m wide zone consists of several dis-
crete cataclastic faults, partly with clay mineral infilling (gouge), and neighboring
zones of dense fracturation. Along this zone the most significant inflows to the
tunnel occur. Discrete inflows in this zone are abundant but oftentimes obscured
by tunnel support systems.
The discharge rates at the individual sampling locations along the tunnel pro-
file range from drop-inflows with mean rates ofO.56-0.85ljmin to more significant
inflows with mean discharge rates up to 33.8 ljmin. Even though small varia-
tions from the mean value of discharge over time can be observed in the manual
measurements at these locations, the automated stations at SI and S2 show no
significant small scale temporal variations of discharge. Figure 2.2 depicts the dis-
charge rates at these stations for the recent period of one year together with the
daily precipitation at the meteorological station GrimseljHospiz (1980 m a.s.l.),
located in the northern vicinity of the research area. Both discharge records show
a steadily increasing trend but no direct response to recent precipitation in this
period. Fluctuations are only minor with respect to the registration resolution
of the gauges at SI and S2 (0.7 and 0.9 l/min respectively). However, the small
but consistent increase of discharge might be linked to the large scale annual
fluctuations of total precipitation. Figure 2.3 shows the annual precipitation at
Grimsel/Hospiz for the period from July 1990 to Juli 2000. It is apparent that
the past three years have been characterized by high total precipitation and tlms
the discharge rates at the stations might be showing a delayed response to these
long-term variations in precipitation.
Total discharge along the whole rv1.5 km tunnel section amounts to a mean
value of 890 ljmin. The variations over time around this average value are again
small (a=18ljmin) and are within the range of the potential measurement error
(24l/min).
Table 2.2 gives an overview of the manually measured in situ parameters for
the individual observation points along the Bedretto-tunnel. The temperature
measurements show only little variation over time at the individual sampling
location and along the tunnel profile, except for a significant negative temperature
anomaly at the locations T9 and to a lesser degree also at SI. Temporal variability
at T9 is the highest (a=0.27 Oe). These two sampling locations are associated with
10
the large fault zone encountered along the gallery (Figure 2.1), where ground
water inflow to the tunnel occurs at high discharge rates (along the whole fault
zone section rv4501/min). Figure 2.4 reproduces this anomaly below the glaciated
area and indicates the potential contribution of cold glacial waters to the ground
water recharge.
The hydrochemical composition of the ground water encountered along the
tunnel profile can mainly be described as dilute, moderately alkaline Na-Ca-
HC0
3
-S0
4
(-F) waters with a mean Ca/Na-ratio of 0.64 (a=0.21; deduced from
meq/l-concentrations) .
Saturation indices show undersaturation with regard to Calcite or sulphate-
minerals as weIl as for Fluorite for all sampIes taken from the tunnel inflows
(mean value for SICalcite=-0.184 a=0.051; SIGypsum=-3.284 a=0.159; SIAnhydrite=
-3.535 a=0.156; SIFluorite=-0.729 a=0.205). These waters have most probable
evolved from the ground water composition as measured at the surface discharges
mainly through the process of Feldspar (Plagioclase) hydrolysis along the flow
paths through the granite body. Petrographic studies of the Rotondo Granite
(Hafner 1958) state a composition of the rock matrix of 25-35% Quarz, 20-
40% Potassium Feldspar, 20-35% Albite, 2-7% Anorthite and 5-10% Biotite
with minor amounts of Garnet, Chlorite, Epidote and Apatite. Local significant
deposits of Pyrite have been reported during the tunnel construction. During
silicate weathering, sodium is released from the Albite-rich granite. This process
results in an alkalinity increase and calcium is most probable lost to the solid
phase by ion-exchange processes on clay minerals, contained in fillings and linings
of fractures and/or fault zones. Precipitation of Calcite is unlikely to occur as
both spring sampIes as weIl as the tunnel inflows are undersaturated with respect
to Calcite.
At the northern end of the Bedretto-tunnel, Ca-Na-HC0
3
-S0
4
(-F) waters
with a mean Ca/Na-ratio of 1.19 (a=0.07) are encountered. Several factors con-
tribute to this changed Ca/Na-ratio. Hafner (1958) reports a decreasing Albite
content within the granite composition towards the margins of the intrusive body.
Therefore, these waters close to the northern boundary of the granite body reflect
this changing geochemical composition of the host rock. Additionally, the degree
of fracturation and especially the number of large fault zones with significant clay
infilling (gouge) is smaller in this part of the tunnel. Less abundance of these
11
features would then imply less potential for ion-exchange with clay-minerals to
remove calcium from solution. Finally, influences from potentially ca1cium-richer
ground waters, originating in the neighboring orthogneiss and in particular in
the amphibolite-wedges within the gneiss (Figure 2.1) are another potential cal-
cium source as indicated by the analysis of spring discharges. Water from the
amphibolite wedges could not be sampled, however studies in the central Swiss
Alps show that similar Ca-rich compositions can be expected to originate from
the amphibolites (Marechal 1995). Because of the observed strong anisotropy of
the gneiss, with steeply dipping schistosities, mainly aligned sub-parallel to the
main orientation of the Gotthard-Massif (ENE-WSW), a pronounced hydraulic
anisotropy within theses rocks seems plausible. This would imply preferential
flowpaths along the strike of the schistosity planes and together with the pre-
ferred orientation of observed hydraulic active fractures and fault zones within
the granite body (ENE-WSW), potential contribution from theses neighboring
formations would most probable originate in the northeast direction of the granite
body (Leckihorn-area; Figure 2.1).
2.4 Environmental isotopes
To investigate potential sourees of ground water recharge and flowpaths as weIl
as the age of the encountered waters, the contents in environmental isotopes
of water sampIes from tunnel inflows were measured, i.e. contents of
18
0, 2H,
tritium (3H) and 34S,
18
0 contents in aqueous sulphate. Additionally,
18
0 and
2H content in precipitation were analyzed. Aqueous sulphate was precipitated
as BaS04 from solution and 34S and
18
0 analyzed by Isotope Ratio Mass Spec-
trometry (IRMS).
18
0 and 2H-content, the latter after zinc reduction to H
2
, in
ground water and precipitation sampIes were determined likewise by IRMS. Tri-
tium values of ground water sampIes were measured by means of enriched Liquid
Scintillation Counter (LSC). Isotope composition is reported as 8-value of the
heavy isotope = (RSample/RStandard -1) x1000 in units of %0, where R=(heavy
isotope/light isotope), e.g. 34Sj32S. Analytical errors in the above 8
18
0, 8
2
H
and 8
34
S determination are O.l %0, 1.0 %oand O.3 %orespectively. Tritium
determinations are associated with an error of O.S TU (all expressed as 20-).
The conventional standards are Vienna-Standard Mean Ocean Water
12
(VSMOW) for (FH and 6
18
0-values and Canyon DiabIo Troilite (CDT) for 6
34
S.
Tritium values (3H) are reported as absolute concentrations in Tritium Units (TU;
1 TU corresponds to 1 tritium atom per 10
18
hydrogen atoms or 0.118 Bq/I).
All isotope measurements were carried out at the GSF Institute of Hydrology
in Neuherberg, Germany.
2.4.1 Tritium
For the analysis of the tritium content of the tunnel inflows two sampling series
were carried out and analyzed (December 1998 and March 1999). Differences
in the tritium content at the individual sampling points between the two series
do not exceed the accuracy of the measurement of 0.8 TU. The mean tritium
concentrations observed at the sampling points along the tunnel profile (Figure
2.4) show values of 7.1-15.1 TU at the northwestern end of the tunnel. Values
of 2.8-3.5 TU are observed towards the southeastern end of the profile below the
glacier. At the southeastern end the mean tritium content lies at 10.8 TU.
Qualitatively, for continental regions, tritium values of 5-15 TU are attributed
to recent recharge 5 to 10 years) while tritium contents of 0.8 to rv4 TU would
represent a mixture between submodern (recharged prior to 1952) and recent
recharge (Clark and Fritz 1997). Two possible scenarios might elucidate the
presence of the low tritium values and the origin of the submodern component.
The sampling sites, showing these low tritium-concentrations (T9, T11 and Sl)
are situated below the valley bottom and might be influenced by upwelling of
old ground water from within the granite body. A plot of tritium content versus
sodium content (Figure 2.5) shows the negative correlation of tritium values with
sodium content. This correlation is typically observed with the presence of older
ground water components, characterized by higher mineralization, i.e. sodium
contents. However, the sodium concentration of the low tritium sampIes is still
small 0.7 meq/l). Other studies in the central Swiss Alps show, that for a causal
relationship between sodium and tritium contents with respect to a source of
submodern ground water, sodium concentrations of at least>1 meq/l are common
(MarechaI1998). With regard to the petrographic analysis ofthe Rotondo granite
(Hafner 1958), the resulting correlation between low tritium values and small but
increased sodium concentrations might merely be caused by coincidence with the
observed spatial variation of Albite content of the rock and thus the increased
13
availability of sodium to go into solution, away from the margins of the granite
body. As the tunnel profile represents a transect from the margin of the body
towards its center, this trend is reproduced in the water composition sampled at
the specific locations.
Another scenario with regard to the influence of the glaciated areas above the
observed inflows to the tunnel is also plausible. Meltwater from the glaciated
areas is derived from a distinct reservoir, which preferentially accumulates win-
ter precipitation with low tritium contents according to the observed seasonal
variation in precipitation. However, the observed mean tritium concentrations in
recent winter precipitation range from 8.4 to 15.8 TU (observations during the
1990's at the stations meteorological stations Guttannen, Grimsel and Meiringen
in the northern vicinity ofthe research area). Therefore, ground water infiltration
from this accumulated recent winter precipitation can only explain the observed
data with lower values down to 2.8 TU, if the residence time of this water is
assumed to exceed the half-life cyc1e of tritium (12.43 years).
Another variant of this latter scenario is regarded to be most plausible. As-
suming that not only recent accumulated winter precipitation with seasonal low
tritium concentrations contribute to the ground water recharge, but also meltwa-
ter from old submodern glacial ice, a source of ground water recharge is present,
which can explain the measured low tritium contents. This source is tapped
by the large fault zone associated with the observation points T9 and 81. On
a plan view projection (Figure 2.1) it can be shown, that the inferred trace of
this structure passes beneath a large portion of the glaciated region which reach
up to altitudes of 3000 m a.s.l. Ground water recharged along strike of this
structure is anticipated to move by topography driven flow and enhanced by the
drainage to the subsurface gallery, along this preferential flowpath towards the
tunnel, whereby preserving the characteristic low tritium values of recharge from
the glacial meltwater of the subglacial drainage system (Tranter et al. 1996). A
similar influence of meltwater from tritium-free glacial ice on the overall tritium
content of glacial discharge is for example also reported by Ambach et al. (1971)
and for the Grimsel area in the northern vicinity of the study area by Keppler
(1996).
The tritium values of 7.1-15.1 TU are in the range of the tritium concentra-
tions in recent winter precipitation. For comparison the tritium concentration
14
in recent summer precipitation are generally higher ranging from 14.7-25.3 TU
(observations during the 1990's at the stations meteorological stations Guttan-
nen, Grimsel and Meiringen in the northern vicinity of the research area). Thus
the observed tritium contents reflect the selectivity (Gat 1981) of ground water
recharge mainly from recent accumulated winter precipitation. Snowcover in the
study area is generally substantial and apart from the high availability of snow
meltwater during the melting period, Rodhe (1998) describes the meltwater in-
put to ground water recharge as very efficient. The snow temperature can not
exceed 0
0
C even if the atmosphere may be warm (and moist). Warm air will
therefore give a small or negative vapor pressure difference between snow and
the atmosphere and thus a small or negative evaporation (i.e. condensation).
Furthermore evaporation from water under the snowpack is negligible due to low
temperatures and lack of turbulent exchange with the atmosphere.
The spread of the encountered tritium data can be attributed to varying flow
velocities (reflected by the different tritium contents) within separate hydraulic
features such as fractures and small scale faults. In this sense, the ground water
system might be viewed as an assemblage of discrete water parcels which move
along defined flow paths (Gat 1981). Similar observations were made by Fontes
et al. (1981) in a comprehensive study in the Mont-Blanc tunnel, where rapid
recharge was indicated by the tritium data with flow velocities of 2;::150 m/a.
2.4.2
34
8/
18
0 in aqueous sulphate
One series of sampIes was collected in November 1999 for the analysis of the
34Sj180-content in aqueous S O ~ - in the ground water encountered in the Bedretto-
tunnel. c5
34
S-values range from 3.35 to 10.24 %0. These values are in the range
of commonly encountered c5
34
S-contents of sulphide minerals in granites (Fig-
ure 2.6). To investigate the origin of the ground water aqueous sulphate more
closely, analysis of its c5
18
0-content is used to obtain additional information, be-
cause oxygen is generally more specifically affected by isotopic fractionation of
redox and phase transition processes than sulphur. Potential origins of the sul-
phate in ground water might be related to anthropogenic sources (e.g. infiltration
of recent sulphate-bearing precipitation or surface waters), oxidation of sulphide-
minerals (e.g. Pyrite), sulphate derived from sedimentary origin (e.g. dissolution
of sulphate-minerals) as also sulphate originating from residual marine formation
15
waters or from residual brines. The latter two potential sources can be neglected
for the situation of the Rotondo-Granite as marine formation waters and basin
brines are not expected to occur. Additional fractionation caused by precipita-
tion is not plausible due to the clear undersaturation of the ground water with
regard to sulphate-minerals such as Gypsum or Anhydrite. This also makes the
potential source of sedimentary origin, i.e. sulphate-minerals, unlikely. However,
contributions of ground water from neighboring formations to the flow system
within the granite might reflect isotopic signals derived from the dissolution of
sulphate-minerals.
A plot of 0
34
S versus 0
18
0 reveals that the isotopic signature of the ground
water lies in the region of sulphate derived from oxidation of reduced inorganic
sulphur compounds (according to Krouse and Mayer (2000); Figure 2.7). A fur-
ther diagnostic tool to investigate the origin of the isotopic signal in the aqueous
sulphate is to consider the relation between the 0180-value in aqueous sulphate
(0180S04) and the 0180-value in H
2
0 (0180H20) of the corresponding sample.
Figure 2.8 depicts possible responses of this relationship according to different
processes. A comparison of the 0180H20/0180S04 signal in the samples from the
Rotondo Granite to earlier published data (Krouse and Van Everdingen 1986)
shows that the Rotondo ground water plots in the region with S O ~ - derived from
oxidation of primary metal sulphides in igneous rocks and with S O ~ - derived
partly from redox reactions (Figure 2.9). The above findings make the oxidation
of sulphide-minerals, i.e. Pyrite, the most plausible scenario as source for the
aqueous S O ~ - . Pyrite deposits have been reported during the tunnel construc-
tion and the 03
4
S signal of the aqueous S O ~ - , which would undergo only little
fractionation during this oxidation process, shows values which are in the range
of reported values for Pyrite (:::;10 %0; Pearson and Rightmire 1980; Balderer et
al. 1987; Balderer et al. 1990).
A detailed plot of 0
18
0 versus 03
4
S in aqueous S O ~ - reveals a trend aligned
to a slope of rvO.6 (Figure 2.10). Such alignments in 03
4
S/0
18
0 data are often
observed (Michelot and Fontes 1987) and have different causes. Basically, such
an alignment might represent the evolution along a precipitation line of sulphate
minerals, along a reduction line or along a mixing line. Precipitation is an un-
likely cause for this alignment as the waters are undersaturated with regards to
sulphate minerals such as Gypsum or Anhydrite. During bacterial reduction,
16
the remaining fraction of aqueous sulphate becomes enriched both in 5
18
0 and
5
34
8. The enrichment factor would thereby be about four (Rafter and Mizutani
1967). Because the slope of the 5
34
8/5
18
0 relationship is much smaIler, mixing
seems the most plausible cause for this alignment. In this interpretation how-
ever, care should be taken though the data values only span a rather narrow
range from which the trend is deduced, both in 5
18
0 and 5
34
8. Furthermore, the
data represent only a singular sampling series. However, it is beneficial for the
understanding of the flow system to investigate potential end-members assuming
that a true mixing line is observed.
Oxidation of Pyrite leads to a fractionation of 5
18
0 with contributions from
both H
2
0 as weIl as molecular oxygen (0
2
). The 5
18
0 in aqueous 8 0 ~ - follows
thereby the equation (Clark and Fritz 1997):
5180S04 = fH20 X (5180H20 + E
18
0S0
4
_H20) + 0.825 x f02 X (5
18
0
02
+
+ E
18
0S0
4
_0
2
) +0.125 x (5180H20 + E
18
0S0
4
_H20)
(2.1)
Lloyd (1968) states the sulphide oxidation by dissolved oxygen to proceed in
two steps (via 8 0 ~ - ) , whereby molecular oxygen and water oxygen are used in a
3:1 ratio. As an approximation, fH20 and f02 were therefore chosen as ~ and ~ ,
respectively. Published fractionation factors Erange from -8.7 to -11.4 %0 for
E
18
0S0
4
_0
2
(Lloyd 1968; Taylor et al. 1984) and from 2-4 %0 for E
18
0S0
4
_H2
0
(Toran and Harris 1989).
Calculation of 5180S04' based on equation 2.1 and inserting 5
18
0
02
=-23.5 %0
(Dole et al. 1954; Horibe et al. 1973) as weIl as 5180H20-values for the individual
sampling location, approximate the observed 5180s04-values only for the data
pairs at the upper end of the correlation plot. This would indicate the oxidation
of Pyrite as a plausible scenario for this end-member of the mixing line. As to the
source of the other end-member it can be proposed, that these lower 5
18
0 and
5
34
8 values are influenced by more complex and possibly multiple redox-processes,
causing additional fractionations (refer to Krouse and Mayer (2000)). As this
5
18
0_ and 5
34
8-depleted end-member is found towards the northwestern end of
tunnel below the core of the massif, more complex reaction/fractionation patterns
along the flow path due to Ionger residence times seem plausible. Alternatively it
can be argued, that ground water contributions from the northern metamorphic
17
gneiss, e.g. with potentially depleted b
34
8 content (compare Figure 2.6 and 2.1),
might be involved in producing the isotopic signal at these locations.
2.4.3 2H/
18
Q in precipitation, glacial meltwater and ground
water
As a furt her means to evaluate the ground water recharge conditions in the study
area, 2H and
18
0 contents in ground waters sampIes were analyzed in compari-
son to b
2
H and b
18
0 values in precipitation and glacial meltwater. Qualitative
observations were reproduced through a simple mathematical transport model
using a convolution/lumped-parameter approach (Maloszewski and Zuber 1996;
Maloszewski and Zuber 1998) to investigate recharge conditions and in particular
the contribution of glacial meltwater to the fiow system. In this approach, an
observed output function, e.g. b
18
0 content over time in the ground water is re-
lated to an observed input function, e.g. b
18
0 content over time in precipitation
by the convolution integral. The type of transport model is chosen by defining a
transit time distribution function and parameterizing this latter function in or-
der to calibrate the theoretical output of the convolution integral to the observed
b
18
0 record.
2.4.3.1 2Rj18Q in precipitation
Data on the b
18
0 and b
2
H content of precipitation were obtained from three me-
teorological stations in the Haslital in the north of the study area (Guttannen,
Meiringen and GrimseljHospiz; 1970-1999) and from five additional stations in
the eastern, western and southern vicinity of the research area (Gtsch, Ander-
matt, Oberwald, Binn, Robiei; 1998-1999). These stations were chosen to cover
the trajectories of potential precipitation to the study area. From this data, the
local meteoric water line (LMWL) for the study area was derived following
b
2
H = (7.65 0.05) x p8Q + (4.7 0.77) (2.2)
Figure 2.11 depicts this LMWL and indicates its relationship to the meteoric
water line (MWL), defined by Craig (1961) as the locus ofthe isotope composition
of world-wide freshwater sources. From the same above data set, the b
18
0-
altitude (z) relationship was derived as:
18
5
18
0 = (-0.0023 0.0003) x z - (9.4 0.5). (2.3)
The above ca1culated 5
18
0-altitude correlation is in good agreement with
published data (Pearson et al. 1991), giving a correlation for the northern central
Swiss Alps of (Figure 2.12)
5
18
0 = (-0.0021 0.0006) x z - (10.6 0.7).
The 5
2
H-altitude (z) relationship was derived as:
52H = (-0.0146 0.0025) x z - (71.46 3.86).
(2.4)
(2.5)
The mean annual deuterium excess d = 52 H - 8 X 5
18
0 in precipitation at the
above meteorological stations ranges from 8.3 %0 at Oberwald in the northern
vicinity of the study area to 13.3 %0 at Robiei in the south of the study area.
Subtle seasonal variations in d can be observed, with deuterium excess values dur-
ing the summer of 7.2-12.1 %0 and during winter of 9-14.2%0 respectively. In
the latter monthly maxima up to 21 %0 are observed. A general explanation for
increased d-values in winter precipitation lies in the type of precipitation during
this period. Solid precipitation such as snow during the winter time shows higher
d-values due to non-equilibrium condensation during the growth of the ice parti-
eIes (Gat 2000; Jouzel and Merlivat 1984). Additionally, Schotterer et al. (1995)
report that winter precipitation in the Alps originating from the Mediterean sea
often show increased d-values.
2.4.3.2 2Hj18Q in glacial meltwater
Glacial meltwater is characterized by a complex signal owing to varying fractions
of snow and icemelt contributing to the overall meltwater composition through-
out the melting period. Furthermore, firnification ofthe accumulated snow cover,
percolating meltwater and liquid precipitation through the glacial body, sublima-
tion and condensation as weIl as melting and recrystallization cause additional
fractionations in the 5
18
0 ofthe ice and snow (Stichler and Schotterer 2000). Un-
fortunately, in the scope of this study no sufficient data was available to clearly
19
constrain these relationships. To investigate the significance of the contribution
of glacial meltwater to ground water recharge a strong simplification was made
for the purpose of the succeeding mathematical model (Maloszewski and Zuber
1996) of 0
18
0 values.
Glacial meltwater is regarded as a combination of snow- and icemelt. At the
onset of the ablation period the contribution from snowmelt dominates while at
later times, melting of glacial ice becomes increasingly impartant (Malard et al.
1999). As each of these two components can be considered as dominant far an
equally long time during the total ablation period, the overall 0
18
0 concentration
of glacial meltwater (c) was derived from the arithmetic mean of 0
18
0 values
in snow and glacial ice respectively (csnow and Cice). From balance equations
Ambach et al. (1971) show that the fraction of ice meltwater in total glacial
discharge may even reach up to 60 %. The 0
18
0 concentration of the snowmelt
(csnow) was approximated from the mean annual value of winter precipitation
(1970-1999) extrapolated to the mean altitude of the glacier. Values for the
concentration in the glacial ice (Cice) were measured in five ice sampIes taken
from the glacier. These few ice sampIes can not be regarded as representative far
the heterogeneous and complex signal originating from the glacial ice (Stichler
and Schotterer 2000). It should rather be considered as an estimate for the
potential contribution of glacial ice to the overall glacial melt component. The
ca1culated estimate for the 0
18
0 concentration in the glacial meltwater amounts
to c =-16.2 %0.
2.4.3.3 2Rj18Q in ground water
Ground water sampIes for 2H and
18
0 determination were collected on a monthly
or bi-monthly basis for aperiod of one year (September 1998 until September
1999) along the rv1.5 km accessible section ofthe unlined Bedretto-tunnel. Eleven
locations along the tunnel profile were sampled (Figure 2.1). The 0
18
0 values
observed in the tunnel show only small variations. The meteoric origin of the
ground water is shown by its alignment to the global meteoric water line (Figure
2.11). Figure 2.4 shows the mean 0
18
0 values measured at the individual sam-
pling points along the tunnel profile. Mean deuterium excess d of the sampled
ground water range from 13.3-14.5 %0. Compared to the d-values in recent precip-
itation at the meteorological stations this may indicate a significant contribution
20
of accumulated winter precipitation with relative higher d-values to ground water
recharge.
Generally, encountered
18
0 values seem to reproduce the altitude effect, i.e.
increase of overburden implies potentially higher recharge altitudes implying de-
pleted
18
0 in precipitation. From equation 2.3 and 2.5 estimated mean recharge
altitudes for the tunnel inflows range from 2600-2800 m a.s.l.
Although Figure 2.1 shows that a straight-forward delineation of potential
recharge areas is not simple due to the complex and steep topography, it is antic-
ipated that recharge and flow through the granite body is preferentially associated
to the ENE-WSW striking discontinuities with a strong vertical component to-
wards the subsurface gallery. This in turn implies that when advancing along the
tunnel profile in southeast direction, i.e. from below the NE-SW trending ridge
towards below the Gerental-valley bottom, the tunnel inflows are recharged from
decreasing altitude zones. This trend is then reflected by the subsurface sampIes.
An intriguing difference from this general trend can be observed at the sam-
pling points T9, Tl! and S1. Here, anomalously negative 5
18
0 values are ob-
served than would otherwise be predicted by the decrease in overburden and thus
in potential recharge altitude. The 5
18
0 depleted values can be attributed to two
factors. These are for one the additional component from the glacial meltwater in
the ground water recharge via the fault zone to the gallery as weIl as an additional
altitude effect, as indicated by numerical simulations of the ground water flow
(Ofterdinger et al. 2001). The 5
18
0 composition of the sampled ground water in
the northern part of the tunnel can be assumed to be derived by recharge from
accumulated winter precipitation and actual summer precipitation during the
recharge period. The ground water sampIes taken along the fault zone however
show the imprint of recharge by the two above components plus the additional
component of glacial ice. Even though this additional component is enriched in
18
0 due to fractionation processes in the transition from snow to ice (Stichler
and Schotterer 2000), the resulting 5
18
0 content is still isotopically lighter than
actual summer precipitation. This additional glacial component will thus shift
the resulting overall 5
18
0 concentration in the ground water recharge from the
glaciated areas towards more negative values.
Additionally, a numerical ground water flow model of the study area (Ofter-
dinger et al. 2001), in which the major fault zone was inc1uded with discrete
21
higher conductive elements, shows that recharge via the fault zone to the tunnel
occurs mainly along the upper slopes of the glaciated areas between Geren- and
Muttenglacier plus to the south along the upper region of the northfacing slope of
the Rotondo summit. This higher recharge altitude compared to the neighboring
sampie locations in the north of the tunnel is also contributing to the depleted
0180-values encountered along the fault zone section of the tunnel.
Figure 2.13 shows the temporal variations in 0
18
0 values for ground water
sampled at observation point T9 as well as the mean signal from the remaining
observation points. The temporal variations at the individual sampling locations
are small (::;0.2 %0, highest along the fault zone section), indicating a lack of
correlation to small scale temporal variations in 0
18
0 of precipitation.
2.4.3.4 Lumped-parameter model
To analyze the relationship between precipitation input and glacial meltwater in
ground water recharge, the observed 0180-values were introduced in a simplified
transport model using a convolutionjlumped-parameter approach (Maloszewski
and Zuber 1996; Maloszewski and Zuber 1998). Even though the 0
18
0 seasonal
variations at the tunnel infiows show only a highly damped response (Figure
2.13), investigation by means of model calculations seem beneficial to asses the
role of precipitation and glacial meltwater in ground water recharge.
In this investigation we applied two approaches to reproduce the observed
0180-values in ground water sampies taken along the major fault zone in the
tunnel. First we introduced the precipitation input into the model to investi-
gate whether observed 0180-values in the tunnel can be reproduced solely by
this precipitation-derived input function. In a second approach we modified the
precipitation input function by adding a glacial meltwater component.
The 0
18
0 input function for precipitation (1970-1999) was derived from the
stations Guttannen, Meiringen and Grimsel (1970-1999) as well as from addi-
tionally analyzed data from the stations Gtsch, Andermatt, Oberwald, Binn
and Robiei (1998-99). The extrapolation to the mean recharge altitude of the
massif above the tunnel was carried out following equation 2.3 for aseries of
potential recharge altitudes of the massif above the tunnel, ranging from min-
imum elevation of 2200 m a.s.l. to maximum elevation of 3000 m a.s.l. The
resulting 0
18
0 signal in precipitation shows large seasonal variations of I"V 14 %0
22
with S
18
0 values ranging from rv-6 %0 during summer down to rv-20 %0 during
winter months. Finally, weighing of the input function by the mean monthly
ground water recharge rates, derived from the hydrological model (Vitvar and
Gurtz 1999) was performed.
The applied code (Maloszewski and Zuber 1996) solves the direct problem,
ca1culating the theoretical output concentration to a given input concentration
(input function) from the convolution integral for a chosen transit time distribu-
tion (parameterized by the mean residence time) under constant flow assumption.
The inverse problem is solved by a trial-and-error procedure with the possibil-
ity to choose between six model-types for the transit time distribution (expo-
nential, piston-flow, linear, dispersion, exponential-piston-flow and linear-piston-
flow). The best fit between the ca1culated output curve and observation data is
judged by an error criterion (standard deviation 0")-
During the modeling process it could be observed that regardless of model-
type and model parameters, the theoretical output curves ca1culated from the
precipitation input function showed enriched S
18
0 values compared to the ob-
served data. Figure 2.13 illustrates this behaviour. This observation holds even
when the extrapolation of the input function is carried out to the maximum po-
tential recharge altitude (with the potentially maximum depleted S
18
0-signal).
This implies that the input function derived from precipitation data alone can
not explain the observed S
18
0-depleted data. This discrepancy indicates that
recharge from a S
18
0-depleted source such as accumulated recent winter precip-
itation and/or glacial meltwater has a significant share in the overall recharge.
Two main scenarios need to be considered to address this discrepancy.
For one, recent depleted winter precipitation is accumulated and stored in the
snowcover during the winter and enters in large amounts the flow system during
the melting season in spring and summer, when most of the recharge to the
ground water flow system occurs. As weighing of the input function is done by
the calibrated recharge rates (Vitvar and Gurtz 1999), this latter observation is
included in the analysis. The S
18
0 concentrations in the input function and thus
in recharge are derived from precipitation. However, this precipitation-derived
input function does not properly represent the S
18
0 input to the flow system
during the melting period. Secondly, an additional source of depleted S
18
0 values
is present in parts of the recharge area in form of the glaciers. This additional
23
source of depleted 6"
18
0 is not accounted for in the precipitation-derived input
function.
Considering these scenarios it seems plausible to modify the input function
to accommodate the 6"
18
0 concentrations of the glacial meltwaters. Therefore in
a second approach we modified the input function to the model by combining
the monthly precipitation data with the glacial meltwater component. In the
numerical ground water flow model (Ofterdinger et al. 2001) the ground water
recharge area to the fault zone section along the tunnel (and thus to sampling
point T9) was delineated through partic1e tracking. As mentioned before this
recharge area comprises parts of the Geren- and Witenwasserenglacier as weIl as
parts of the northfacing slope of the Rotondo summit. The calibrated hydrological
model (Vitvar and Gurtz 1999) resolves apart from the monthly ground water
recharge rate in this area also the fraction of meltwater and actual precipitation
participating in the corresponding ground water recharge. The 6"
18
0 signal of the
modified input function was derived by combining the 6"
18
0 signal in precipitation
and the 6"
18
0 signal in the meltwater (c) according to their contribution to the
monthly ground water recharge.
From the above available model-types, a best fit in terms of shape of the
output function as weIl as in error criterion was achieved by applying the expo-
nential model combined with a piston flow component (EPM). The applied ratio
of total volume to the volume with the exponential distribution of transit times,
Tj, was hereby 1.2. The necessity to apply a combined model in order to achieve a
good fit as in this case is most probably linked to the nature of the granitic rocks.
These provide, due to zones of dense fracturation and faults as weIl as more intact
massive rock, different flow components with different transit time distributions.
As the modeled 6"
18
0 signal (sampling location T9) is associated with the major
fault zone with dense fracturation, it is not surprising that a combined model
leads to the best fits.
The best fit of the EPM-model to the observed 6"
18
0 data (Figure 2.13) yields
a mean residence time of rv14 months with an error criterion of Cf = 0.0096 %0.
The lack of strong variations in the observed 6"
18
0 signal in the ground water
despite the short residence time is mainly attributable to the large fraction of
glacial meltwater contributing to the ground water recharge (
rv
70 % of annual
recharge) compared to liquid summer precipitation.
24
2.5 Discussion and Conclusion
Several approaches have been shown to give information on the origin of the
ground water eneountered in the Rotondo granite and espeeially on the eonditions
of recharge, in partieular the influenee of the glaciated regions in the study area.
While the interpretations of the individual investigated parameters offer already
potential answers to the above questions, a more eompelling scenario ean be
dedueed by eombining the results of the individual investigation.
In general, all investigated parameters show, that at the tunnel elevation, only
little temporal variations ean be observed. The measured ()ZH/6
18
0 signal in
ground water shows its meteorie origin (Figure 2.11). Hydroehemieal analysis of
the tunnel inflows shows, that their eomposition ean be explained as an evolution
along the flow path within the granite body. Flow is hereby assumed to be
preferentially orientated along the steeply dipping fraetures and faults, following
an overall ENE-WSW strike. Divergenee of Ca/Na ratios towards the northern
margin of the granite might weIl be explained by petrographie variations within
the granite but might also be influeneed by ground water eontributions from
neighboring Ca-rieher formations, i.e. from the orthogneis and espeeially the
amphibolite wedges.
The analysis of the 6
34
S and 6
18
0 in aqueous S O ~ - underlines that the en-
eountered ground water is derived from within the granite body, where aqueous
S O ~ - is mainly derived from oxidation of Pyrite. The alignment of the 6
34
S/ 6
18
0
data along a slope of ",0.6 however indieates, that redox-proeesses and more
eomplex kinetie fractionations might influenee the waters eneountered below the
eore of the massif. Under this assumption the alignment of the 6
34
S/6
18
0 data
represents a mixing of ground water eomponents from within the granite body.
Contributions of ground water originating in the northern neighboring gneiss
would represent an alternative explanation for this mixing line.
Tritium, temperature and 6
18
0 eontent measured along the tunnel profile
(Figure 2.4) show all anomalous behaviour below the glaeiated areas. As this
seetion is at the same time situated approximately below the valley bottom, two
potential explanations for these observations need to be addressed.
Contribution of old ground water eomponents upwelling below the valley bot-
tom would be assoeiated with low tritium contents and potentially depleted 6
18
0
values due to low-temperature roek-water interactions. Both of this would be
25
in agreement with the observation even though 6"
18
0 fractionation would be ex-
pected to be stronger. The tritium/sodium correlation observed is no compelling
indicator for the contribution of older ground water components because sodium
concentrations are still very small and the correlation is most probably only a
spatial coincidence and no causal relationship. The conductivity measurements
and hydrochemical analysis of the ground water give also no evidence of old
ground water components with potentially higher mineralization. Furthermore
the negative temperature anomaly is not explained by this scenario.
A second explanation for the observed data would be a significant contri-
bution of cold submodern glacial meltwater from the overlaying glaciers with
low tritium concentrations and depleted 6"
18
0 values. Model calculations under-
line the contribution of this 6"
18
0 depleted source to ground water recharge and
even though the absolute figures of this model should be treated with care be-
cause of the simplified model assumption, the significance of this contribution is
demonstrated. The large fault zone associated with the inflows where the above
anomaly is observed is providing hereby a preferential flow path. Due to the
sub-vertical ENE-WSW strike of this structure, a potentially large glaciated area
with topographic elevation up to 3000 m a.s.l. (Figure 2.1) is tapped. Ground
water recharged along strike of this structure in the glaciated areas is believed to
move by topography driven flow and enhanced by the drainage to the subsurface
gallery along this preferential flowpath towards the tunnel, whereby preserving
the glacial signature.
On the whole, the above investigations suggest that the ground water flow sys-
tem within the Rotondo Granite is recharged by meteoric waters with significant
contribution of glacial meltwater and meltwater from accumulated recent winter
precipitation.Recharge contributions from the glaciated areas is hereby enhanced
by the orientation of the water conducting fracture sets and fault zones.
Temporal variations in the observed parameters of the ground water are small.
Cross-flow from neighboring formations is minor and is, if at all present, mainly
restricted to the northern boundary of the granite body. The ground water be-
comes moderately mineralized along the flow path within the granite, mainly
through the process of silicate weathering (Plagioc1ase hydrolysis). Model calcu-
lations approximate the mean residence times of the encountered ground water
along the fault zone section in the gallery of rv1-1.5 years and underline the sig-
26
nificant contribution of glacial meltwater to ground water recharge. Even though
the absolute figure of the mean residence time should be regarded as approxima-
tions due to the involved simplifications, it indicates rather rapid recharge from
the glaciated areas via the fault zone to the gaIlery.
Acknowledgements
Precipitation sampIes from the meteorological stations at Gtsch, Andermatt,
Oberwald, Binn and Robiei were kindly taken by the station supervisors. Data
from the stations Guttannen, Meiringen and Grimsel where obtained from the
Schweizerische Meteorologische Anstalt-SMA, Zrich and the Landeshydrologie,
Bern as weIl as extracted from the GNIP database provided by the IAEA, Vienna.
Access to the Furka-basetunnel and the Bedretto-tunnel and logistic help was
kindly provided by the Furka-Oberalp-Bahn. This research is funded by the
ETH research fund.
27
Table 2.1: Mean orientation of fracture sets and corresponding mean normal
fracture frequency A (Priest 1993) from surface and tunnel outcrops.
Number of set Mean orientation
Asur face Atunnel
strike/dip
[m-l] [m-l]
1 049/75 SE 2.1 0.5
2 080/83 SE 1.5 0.2
3 140/86 SW 0.7 2.6
4 170/79 SW 0.04 0.3
28
Table 2.2: Mean in situ parameters (discharge Q, temperature, conductivity 1'1"
pR, total dissolved solids TDS and dissolved oxygen O
2
) at individual sampling
locations along the tunnel as wen as mean in situ parameters for spring discharges
from the Rotondo Granite (,R) and the neighboring orthogneis (G; refer to Figure
2.1). Standard deviation in brackets as indicator for variability over time.
Location Q Temperature K, pR TDS O
2
[l/min]
[0G]
[1L
8
/
cm
] [meq/l] [mg/l]
Tl 0.6 19.3 83.1 8.9 1.64 2.7
(0.11) (0.2) (1.4) (0.2) (0.07) (0.2)
T2 26.9 19.7 88.1 9.0 1.73 2.3
(1.22) (0.1) (1.4) (0.2) (0.07) (0.2)
T3 0.6 19.9 71.8 9.1 1.37 3.0
(0.03) (0.2) (1.6) (0.1) (0.10) (0.1)
T4 0.8 19.3 94 8.9 1.89 1.1
(0.07) (0.2) (1.6) (0.1) (0.12) (0.3)
T5 2.1 19.7 89.1 9.1 1.71 2.4
(0.23) (0.1) (1.5) (0.1) (0.10) (0.3)
T6 1.1 19.6 89.4 9.1 1.74 1.9
(0.10) (0.1) (1.4) (0.1) (0.05) (0.2)
T7 6.6 19.0 104 9.1 1.99 1.8
(0.85) (0.1) (1. 7) (0.1) (0.14) (0.3)
T9 10.3 15.4 91.5 9.2 1.76 1.2
(0.65) (0.27) (1.9) (0.1) (0.12) (0.2)
T11 3.0 19.6 115.1 9.3 2.14 0.3
(0.27) (0.05) (1. 7) (0.1) (0.14) (0.1)
81 18.4 17.7 105.5 9.2 2.10 2.0
(0.47) (0.1) (2.2) (0.1) (0.10) (0.2)
82 30.5 18.9 79.2 8.9 1.54 2.9
(1.91) (0.04) (1.4) (0.1) (0.10) (0.3)
Q (,R) 7.9 16.0 7.2 0.35 8.1
(1.6) (1.5) (0.3) (0.05) (0.8)
Q (G) 6.5 64.7 7.3 1.27 6.7
(0.6) (2.5) (0.3) (0.05) (0.2)
29
.,. __ .,. .,. Geologieal boundary
__.. Major fault zone (partly inferred)
yR Rotondo-Granite
G Orthogneiss
am Amphibolite
Spring diseharge
elevation eontour interval =35 meter
..3070 Topographie referenee with
elevation in meters a.s.1.
le Leekihorn
ro Rotondo
Glaeiated area
wg Witenwasseren-glacier
gg Geren-glacier
Figure 2.1: Enlarged plan view of the research area. Indicated are the projected
tunnel traces together with the location of the subsurface observation points,
geologieal boundaries as wen as the trace of the major fault zone intersecting the
Bedretto-tunnel. Furthermore locations of spring discharges and glaciated areas
are marked.
30
JulOO AprOO JanOO
Date [mmyy]
Oct99
45.-----------,-----------,,---------,-----------,
40
1l, 25
(ij
13 20 L ~ " ' - ' - ' - ' ~ ~ - - - - - - - - - 7 " ' " - - - - ~ - ~ ~ ~ - ~
'" Ci 15
10
5'----------'-------------.J'-------------'-------------'
Jul99
40.-----------,-----------;,-------------,-----------,
_ Daily precipitaion at
Gnmsel/Hospiz
E 30
oS
c
.2 20
J'l
ii
U
~
a. 10
Oct99 JanOO
Date [mmyy]
AprOO JulOO
Figure 2.2: Record of automated discharge measurements 81 and 82 in the
Bedretto-tunnel together with daily precipitation record at the meteorological
station Grimselj Hospiz (1980 m a.s.l).
3000 .--------r---r--...,...--.,....--.-------r----r---r--.,....----,
2500
E
.s 2000
c
o
2 1500
"0..
"0
~ 1000
0..
500
o
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Year
Figure 2.3: Record of annual precipitation (1990-2000; summed for period July-
July) at the meteorological station GrimseljHospiz (1980 m a.s.l.).
31
SE NW
3200
0
500. m
,.......,
3000
sampled inflow
u:i
co
.s
2800
Q)
"0
2600
:::l
-
:;::;
2400
Tunnel collapsed T9 81 T11 T2
1480
__________ \ ~ __ __.... _...__ .!"Y__ "! _... _ %%! __
22.5
~
L
20.0

Q)

I-
:::J
......
~
17.5
Q)
Cl.
E
15.0

~
12.5
16
llil
f
S'
12 f ff
t::..
I
I
E 8
:::l
I :;::;
iE:
4
11
f
0
-14.6
[Q]
fl
,.......,
ff
0
cf( -14.8
0
I ff
<Xl
'i:l
-15.0
-15.2
Figure 2.4: Cross-section along the profile of the Bedretto-tunnel with locations
of sampled inflows; a) mean temperature, b) mean tritium concentration and c)
mean 6"
18
0 content. Analytical error is indicated.
32
1.0 , . . . . - - - - , . . . - - - - ~ - - - _ _ r - - - _ _ _ ,
r"=0.80
0.8
g0.6
.s
E
:6 0.4
o
Cf)
0.2
limit of
0- 95%-confidence interval
CII>'"'-. I
o "
' ~ ~
, """-
, """-
-s.... 'c9-
~ " 0
o 0 0,
0'0 ......
o ""
'"'-.
"
slope=-0.03
20 15 10
Tritium [TU]
5
0.0 L...- .l...- ...L.- ....L- ...J
o
Figure 2.5: Plot of Tritium versus sodium content for sampIes from the Bedretto-
tunnel.
'--__....JI Evaporite sulphate
oOcean water
Sedimentary rocks
Metamorphic rocks
c=:JGranite rocks
oBasaltic rocks
DRotondo
I I I I I I I I I I
50 40 30 20 10 0 -10 -20 -30 -40
34
S [%0]
Figure 2.6: 0
34
8 values of some geologically important materials in comparison
to encountered range in the Rotondo Granite (modified after Hoefs (1997)).
33
20
-10
atmos-
pheric
evaporites
depo-
sition
-
.
sulphate derived fram "
oxidation of reduced
.
.
.
inorganic compunds
.'

-30 -15 o 15
34
Sso, [%01
30
Figure 2.7: Commonly observed ranges of 5
34
8 and 5
18
0 values for sulphates in
relation to measurements from the Rotondo Granite (.) (modified after Krouse
and Mayer 2000).
'<t
o
cn
o
CD
'Zo
REDUCTION
W/O
EXCHANGE
EVAPORATION
Figure 2.8: Effects of various processes on the 5
18
0 values of 8 0 ~ - and H
2
0
(after Krouse and Everdingen 1986).
34
20
10
o
J
o
c3
(J)
o
<Xl
Zo -10
-20
ME
HzO:: 0 % I __-i':-
._e .
---. ~
r p ~
e_e
e a
b
... c
.d
e
o f
10 o -10
18
0H,O [%01
-20
-30 -!L----.....,------..,.----r------r--1
-30
Figure 2.9: 0180H20 versus 0180S04 values from published data in comparison to
data from the Rotondo Granite. a) dissolved marine sulphates; b) S O ~ - derived
from oxidation ofmetal sulphides; c) Gypsum from H
2
S0
4
fallout; d) S O ~ - partly
derived from redox reactions; e) Laboratory experiments from Lylod (1967; L),
Schwarcz and Cortecci (1974; S) and Taylor et al. (1984; T); marine evaporites
(ME). SampIes from Bohemia (B), Canada (C), Italy (I), Kuwait (K), Poland (P),
USA (U) and sampIes from the Rotondo granite (f,R). Lines refer to the percent
water in the oxidation reaction (modified after van Everdingen and Krouse (1985),
Krouse and van Everdingen (1986)).
35
12
r2=0.79
0/
/0
/
slope=0.62
0
8
/
~
/ 0
,.....0
/0
0
/'
o 0/ (f) /'
Cf)
/ \
"
'" 00
4 / limit of 95%-confidence
0
/
interval
/
0
-8 -4 0 4 8
S180804 [%0]
Figure 2.10: Detailed plot of 6180S04 versus 6348s04 in aqueous sulphate in sam-
pIes from the Bedretto-tunnel.
0
x precipitation h
o groundwater ~
x ~ \
-50
LMWL
~
~
'I' -1 00
N
C>O
-150
2
H=8
x
18
O+10
V (MWL)
~
r2LMWL= 0.987
-200
-30 -10 0
18
0 [%0]
Figure 2.11: Local meteoric water line based on 6
2
H/6
18
0 data from meteorolog-
ical stations together with the meteoric water line as defined by Craig (1961).
Indicated is also 6
2
Hjb'
18
0 data from ground water sampies collected in the
Bedretto-tunnel.
36
-10 r--........--..------.-------.---------,
r
2
=O .916
1 Meiringen
2 Guttannen
3 Sinn
4 Andermatt
5 Oberwald
6 Grimsel
7 Guetsch
8 Robiei
\
\
\
" limit of
"" "95%-confidence interval
............ "
" \
" \
"" \
",,2
0
3"
" ,
"",,5 4 '" 08
, '
, '"
, "
" 6
0
.........
, " ....
\ "
\ ""
\ "
\ ""
\ "
\ "
\
\
"
7
0
\
,
\
\
-16 ' : - - - - - - ' - - - - - - . . . . - . . . , J ' - - - ~ \ ~ - - - ' ......
o 1000
Altitude
-11
-12
-15
-14
,........,
o
:::R
o
.......... -13
o
ro
Figure 2.12: Relationship between altitude and mean annual 6
18
0 content m
precipitation at meteorological stations in the vicinity of the research area.
37
T9
GWmean
Precipitation only;
mrt =14 months
\
- - ............
'"
Precipitation only;
mrt = 72 months
r\
I \
I \
[TI I \
I \
I \ ~
\
\
\
\
\
\
\
\
\
\ /
- -- -- -\ L
.... -
\ -/
\/
-13.4
-13.2
-13.0
-13.6
-13.8
-12.6
-12.4
-12.8
Figure 2.13: Observed 6
18
0 data at the stations T9 and ca1culated mean signal
ofremaining ground water sampling stations (GWmean). Indicated is the output
function for lumped-parameter modeling representing the best fit to the observed
data when the precipitation-derived input function is modified by an additional
meltwater component (line3). Line 1 illustrates the output function far solely
precipitation-derived input function far identical model parameters of the best
fit model. Even for potential large residence times (line 2), the output function
from the precipitation-derived input function remains enriched compared to the
observed data. (mrt = mean residence time)
3
Hydraulic subsurface
measurements and hydrodynamic
modeling as indicators for ground
water flow systems in the
Rotondo Granite
u. s. Ofterdinger, Ph. Renard & s. Loew
Submitted to
Ground Water
38
39
3.1 Introduction
High alpine catchments in fractured crystalline rocks present a great challenge
when assessing the large scale ground water flow systems. Regional flow sys-
tems are strongly influenced by topography and basin scale (T6th 1963; T6th
1984; Zijl 1999). Furthermore, the geology (i.e. permeability and thermal con-
ductivity), c1imatic conditions (available infiltration and surface temperatures)
and regional heat flux are governing factors on regional flow systems (Forster
and Smith 1988a, b). Structural elements such as fault zones also strongly gov-
ern the behaviour of these systems (Forster and Evans 1991; L6pez and Smith
1995). Various conceptual approaches have been developed in the past decades to
describe and model the ground water flow through fractured rock masses (NRC
1996) ranging from equivalent continuum models (Carrera et al. 1990; Bear 1993)
to discrete fracture network simulation models (Cacas et al. 1990a-c; Dverstorp
and Andersson 1989). Numerous case-studies have been performed in fractured
crystalline rock in the framework of the safety assessments for nuc1ear waste
repositories (Kimmermeier et al. 1985; Herbert et al. 1991; Neretnieks 1993)
investigating flow and transport phenomena on the small scale as well as on a
regional scale (Voborny et al. 1991; Voborny et al. 1994). Observations in sub-
surface galleries have been used to investigate ground water flow systems within
fractured crystalline rocks on a regional scale (Loew et al. 1996; Kitterod et al.
2000).
In high mountainous catchments, recharge conditions have strong spatial vari-
ability (Balek 1988; Lerner et al. 1990; Kattelmann and EIder 1991), depending
on numerous factors such as geology/soil cover, topography/landform and vege-
tation/landuse. The impact of varying recharge conditions and specific recharge
sources such as snowpack and glaciers (Martinec et al. 1982) in mountainous re-
gions has been investigated mainly with the focus on shallow ground water flow
systems and characteristic signatures of recharge sources were shown in terms
of environmental isotope content and hydrodynamic response (Flerchinger et al.
1992; Ward et al. 1999; Abbott et al. 2000).
To understand large scale flow systems in crystalline rocks of high mountain-
ous terrain a comprehensive study has been performed in the Rotondo Massif of
the Central Swiss Alps. Various types of measurements were carried out at the
40
terrain surface and in subsurface galleries. These range from the analysis of the
hydrochemical and environmental isotope composition of groundwater and pre-
cipitation (Ofterdinger et al. 2000) to the acquisition of data on the structural
geology and the measurement of hydrogeological parameters of the crystalline
bedrocks. To improve our overall understanding of the ground water flow sys-
tems, numerical simulations following a deterministic continuum approach were
carried out. spatially distributed recharge rates were extracted from the results
of a hydrological model and applied as upper boundary condition for the ground
water flow simulations. Key questions addressed with the ground water flow
model are the influence of spatially distributed recharge on the ground water
flow field and the position of the free water table, the impact of subsurface gal-
leries and the influence of a large scale fault zone on the regional flow system
in this topographically complex terrain. The simulations are additionally aimed
at providing a better understanding of the environmental isotope data and to
test our previously derived interpretations of these (Ofterdinger et al. 2000). In
the following we first introduce the geological and hydrogeological setting of the
study area. Then the available information on hydrogeologieal parameters are
illustrated and our conceptual model is described before details of the modeling
approach and results are given. Conc1uding the discussion of the model results
a comparison to the findings from a previous study of the environmental isotope
composition of the ground water (Ofterdinger et al. 2000) in the study area is
drawn.
3.2 Geological and hydrogeological setting
3.2.1 Major geological and hydrogeological units
The research area is situated in the western Gotthard-Massif of the Swiss Alps.
Geologically it consists mainly of the late Hercynian Rotondo granite and or-
thogneisses of early Palaeozoic age (Figure 3.1). The topographie elevation of
the granite body ranges from 1800 to 3200 m a.s.l (Figure 3.2).
On the northern margin ofthe study area the Furka-basetunnel passes through
the granite body at an elevation of rv1490 m a.s.l. (Figure 3.1). An abandoned
and unlined support segment to this basetunnel, the Bedretto-tunnel, is further-
more passing through the granite body in northwest-southeast direction. Due
41
to the partial collapse of the tunnel, only the northern I"V 1.5 km are still ac-
cessible. Figure 3.2 shows in particular that this latter tunnel segment is also
passing beneath glaciated areas of the study domain. It offers the opportunity
for direct subsurface observations and was used for sampling of ground water and
monitoring of discharges and in situ parameters.
The focus of this investigation is targeted on the ground water flow systems
within the granite body and its immediately adjacent formations. For the model-
ing purpose the domain was however extended and includes additional geological
units (Figure 3.1).
From north to south the geological units are the Hercynian Aar-Granite (A)
and the Southern Gneiss Zone (Ak) of the Aar-Massif. Between these units of
the Aar-Massif and the units of the Gotthard-Massif to the south, the meta-
morphic sediments of the Permian-Jurrassic Urseren-Gavera-Zone (UGZ) can be
found. Further to the south, the Cambrian-Palaeozoic Paragneisses (PN) and
Orthogneisses (0) are followed by the Southern Paragneisses (Pr) within the
Gotthard-Massif. Along the southern margin of the model domain follow the
Triassic-Jurassic Ultra-Helvetic sediments (M) of the Gotthard-Massif. Addi-
tional to the Rotondo Granite, furt her Hercynian granite bodies can be distin-
guished within the domain. These are the Gamsboden/Cacciola-Granite (C) and
the Fibbia-Granite (F). Throughout the area of the Gotthard- and Aar-Massif,
alpine faulting and foliation can be observed (Labhart 1999). The former are
often following pre-existing structures such as lithological boundaries, intrusive
contacts or dykes and are found today mainly with steep to subvertical incli-
nations. Foliation is selective and less developed within the Variscian granites.
These tectonic elements show mainly a NE-SW or E-W strike orientation with
steep or subvertical inclination.
The geological boundaries indicated in Figure 3.1 are simplified for the model-
ing purpose and reflect the geological make-up of the domain on the massif-scale.
The boundaries are extrapolated vertically with depth. The geological record
of the Furka-basetunnel and the Bedretto-Tunnel (Keller and Schneider 1982;
Schneider 1985b) show that for the Rotondo area this simplification is valid due
to the subvertical dip of the geological boundaries. Also in the rest of the model
domain mainly steep to subverticallithological boundaries are observed (Labhart
1999).
42
3.2.2 Structural geology
The Rotondo granite shows varying degrees of fracturation and faulting through-
out the investigation domain, changing both in lateral and vertical directions.
Data on fracture orientation and fracture frequency (A) were gathered by means
of scanline surveys (Priest 1993) on surface outcrops and along the unlined tunnel
section of the Bedretto-tunnel. 23 scanlines with lengths of 10-20 m were mea-
sured. Additionally, geological tunnel records (Schneider 1985a) for the sections
within the Rotondo granite as weIl as aborehole televiewer log from a research
borehole (see below) were analysed leading to a database with 1174 fracture
measurements.
From this database, four major fracture sets can be deduced (Table 3.1),
where especially set 1 and 2 are dominant at the ground surface. The NW-SE
striking fracture sets (3 and 4) are more abundant along the tunnel profile and are
believed to be influenced by the tunnel construction, i.e. induced or reactivated
by local stress amplifications.
In addition to the steeply dipping sets described above, flat dipping fractures
can be identified at the terrain surface. These fractures generally strike sub-
parallel to the valley axes with dip directions towards the valley bottom, most
probably representing quaternary stress release joints. Occasionally, concentrical
jointing can also be observed within the granite at ground surface. In addition to
the fracturation, steeply dipping fault zones can be identified as important larger
features, both at surface and along the tunnel profile. They are mostly orientated
sub-parallel or at acute angles to the general trend of the alpine structures of
the Gotthard-Massif (ENE-WSW). It can be observed that both on the terrain
surface and in the subsurface, fractures of set 1 and 2 as weIl as sub-parallel
orientated fault zones are the dominant water conducting features amongst the
observed structural features (Ofterdinger 2001).
One of the dominant steeply dipping fault zones in the research area is striking
rv 75 and follows the eastern part of the Gerental-valley (Figure 3.2) intersecting
the Bedretto-Tunnel in the subsurface, where it is associated with major inflows.
While the fault zone on the terrain surface is often obscured in the valley floor by
scree deposits, it is consistently apparent through dense fracturation along the
northwestern valley slopes and in the NW-SE trending mountain ridges to the
east of the Bedretto-Tunnel. In the subsurface, the structure of the fault zone
43
can be directly studied along its intersection with the Bedretto-Tunnel.
Fault zones are conceptually subdivided into fault core with fine grained gouge
material and adjacent fractured damage zone towards the surrounding protolith
(Smith et al. 1990; Caine et al. 1996). The observed approximately 120 m
wide zone described in this paper comprises several discrete fault zones of cm-dm
width. While the damage zones of these faults are pronounced and characterized
by dense fracturation, only few lenses of fine grained gouge material occur within
these discrete faults. In the following, this 120 m wide zone will be referred to
as 'fault zone' in the conceptual context of this study, even though this term
originally refers to an indivdual structure of the above described succession of
fault core, damage zone and protolith. For a detailed documentation of this zone
and its structural elements refer to Ltzenkirchen and Loew (2001). Previous
studies (Schneider 1976) have traced this fault zone, which caused substantial
problems during tunnel excavation (Keller and Schneider 1982), on the surface
along the Gerental valley and further to the east to an overall length of more
than 10 km. As this interpretation is little constrained by direct observations
(the fault zone to the east is obscured by glaciated areas and scree slopes), a
conservative approximation of the modeled fault zone geometry was chosen with
a length of rv4.5 km. The fault zone width was approximated to be constantly
120 m, although it is known that generally the width of fault zones is highly
variable and heterogeneous (Smith et al. 1990) both lateral and with depth. Yet,
fault zones with similar widths at surface and in depth have also been observed
(Raven 1977; Wallace and Morris 1979).
3.2.3 IsotopesjChemistry
In a previous publication (Ofterdinger et al. 2000) the composition of the en-
countered ground waters from the Rotondo granite was studied with regard to
environmental isotopes content and hydrochemistry. The analysis of the 34Sj180
content in aqueous S O ~ - of sampled ground water from the Bedretto-tunnel re-
vealed that the encountered ground water is derived primarily from within the
granite body. A potential contribution from the northern neighboring gneisses
might serve as an explanation for the alignment of 34Sj180-data along a mixing
line with slope rvO.6. Varying Ca/Na-ratios along the tunnel profile indicate the
44
possibility of ground water contributions from the neighboring gneisses (0) and
in particular from local amphibolite wedges.
The 0
18
0 measurements of ground water sampIes underline their meteoric
origin and indicate major contributions from accumulated winter precipitation to
ground water recharge. Especially along the above described fault zone, tritium,
0
18
0 and temperature data imply a significant contribution of glacial meltwater
to ground water recharge. Measured 0
18
0 reflect an altitude effect, i.e. higher
recharge altitudes towards the northern end of the Bedretto-tunnel. Estimated
recharge altitudes lie in the range of 2600-2800 m a.s.l.
Along the whole Bedretto-tunnel no indications for upwelling brines or con-
vection driven flow are observed. Estimated mean residence times from lumped-
parameter modeling of the 0180-values in ground water sampIes taken along the
fault zone section are 1 to 1.5 years.
3.2.4 Spring line
Few c1ear spring discharges from the granite bedrock can be observed in the
Rotondo area. Their characteristic chemical composition is discussed in Ofter-
dinger et al. (2000). In the upper Gerental-valley spring occurrences generally lie
at an elevation of 2200-2300 m a.s.l, approximately 100-200 m above the valley
floor (Figure 3.2). Even though these spring discharges were observed to orig-
inate from open fractures in the granite bedrock or associated to fault zones,
the question whether the altitude of these springs truly marks the intersection
of the water table with topography needs to be considered with care, as near
surface flow, enhanced by flat dipping stress-release joints in the decompressed
zone, might mask the true altitude of the water table along the slopes.
3.3 Hydrogeological parameters
3.3.1 Hydraulic heads
For the monitoring of hydraulic pressures within the Rotondo granite a shallow
dipping I'.J 100 m long research borehole was drilled from the tunnel gallery (Of-
terdinger 1999). The borehole is situated at the intersection of Furka-basetunnel
and Bedretto-tunnel below the Saashrner-ridge with an approximate overbur-
45
den of 1200 m and a plunge direction of 165
0
(30
0
deviation from the axis of the
Bedretto-tunnel). The orientation was chosen to intersect the prevailing water
conducting steeply dipping fracture sets within the moderately fractured gran-
ite (ArvO.2-O.5m-1) at dose to right angles. Two borehole intervals have been
permanently packed of by grout injection to monitor the hydraulic pressures.
Monitoring intervals are 99.2-86.0 m (Interval 1) and 50.0-66.0 m (Interval 2)
with perpendicular distances to the axis of the Bedretto-Tunnel of rv43 m and
rv29 m respectively. Previous borehole logging revealed inflows along the inter-
vals of rv4ljmin (Interval 1) and rv16 ljmin (Interval 2) associated mainly with
fractures of set 1 and 2. Interval pressures are monitored via 1" tubings at the
borehole mouth. Monitoring started in July 1999 and is currently ongoing. After
a preliminary phase the final sensor-setup with piezoresistive transmitters was
installed in October 1999 (Type PA-53, Keller eH).
Measured pressures are in the order of 50 bars for Interval 1 and 40 bars
for Interval 2 (Figure 3.3). The recorded signal in Interval 1 shows annual fluc-
tuations in the order of 1-1.5 bar with maximum pressures in November and
minimum pressures in June. Whereas a similar pattern can be observed for 1999
in Interval 2 the pressure rise towards autumn 2000 is less dear. As calibration
target for the individual intervals, the average hydraulic pressure measured for
the period from 1. November 1999 to 1. November 2000 was chosen (Table 3.5).
With respect to the modeling effort, these head measurements are not sufficient
to fully constrain a regional model as they represent only a local areal observa-
tion. Their purpose is mainly to investigate the long-term transient behaviour of
a moderately fractured section of the Rotondo granite.
3.3.2 Ground water fluxes
Measurements of spatially integrated inflow rates to the subsurface galleries were
carried out on a monthly/bi-monthly basis. In the galleries, inflowing ground
water is captured in a basal drain and drained towards the tunnel portals. Inflows
to the Furka-basetunnel east of kilometer 11.5 (Figure 3.4) is drained to the
Realp-Portal while the remaining section is drained to the Oberwald-Portal. No
drainage water from the Furka-basetunnel is passing through the Bedretto-tunnel.
Ground water inflows along this tunnel section are drained to the southern Ronco-
Portal.
46
Measurements of integral flow rates along particular tunnel sections were car-
ried out in the basal drain of the tunnels with depth-integrated flow velocity
measurements (flow-anemometer). Discharge was then ca1culated according to
IS074S:1997(E) (ISO 1997). In the Furka-basetunnel measurement locations
were chosen to characterize the inflow along the intersected section of the Ro-
tondo granite as weIl as the total flow rates at the tunnel portals in Oberwald
and Realp. In the Bedretto-tunnel, measurement locations were chosen to char-
acterize the distribution of inflow rates along the tunnel profile and especially
along the intersected fault zone. Measurements were recorded from August 1995
to January 2000 in the Furka-basetunnel and from November 1999 to November
2000 in the Bedretto-tunnel. Table 3.2 shows that the measured inflow rates show
only small temporal variability, usually within the range of measurement errors.
3.3.3 Hydraulic conductivity
Several previous studies have investigated the potential range of permeabilities
of fractured crystalline rocks both on the laboratory as weIl as on the field scale
(Brace 19S0; Brace 19S4; Clauser 1992) and in particular the permeability dis-
tribution within fault zones and their adjacent crystalline formations (Davison
and Kozak 19S5; Evans et al. 1997; Scholz and Anders 1994; Forster et al. 1994;
Caine and Forster 1999). Commonly a wide range of values are found spanning
several orders of magnitude depending on the degree of deformation and the sam-
pling position within the fault zone or neighboring crystalline rock. Accordingly,
a range of anisotropy ratios for fault zones are stated, describing the oftentimes
observed enhanced fault-parallel flow in contrast to reduced fault-normal flow.
A first approximation of the hydraulic conductivities of the more intact sec-
tions of the Rotondo granite was deduced from hydraulic tests in the research
borehole (constant-rate tests). Four packer tests with increasing test interval
length (7-35.5 m) were carried out starting from the bottom ofthe borehole. Ap-
parent hydraulic conductivities (referenced to intervallength) range from 6.0 x 10-
9
to 2.9x10-
8
m/s with a geometric mean of 1.1x10-
8
m/s.
On the basis of the measured inflow rates to the Bedretto-tunnel and assuming
steady-state conditions, hydraulic conductivities for the fault zone section and the
remaining granite section along the tunnel profile as weIl as for the whole profile
length were approximated using the analytical solution described by Goodman
(3.1)
47
(1965) based on the assumption of a linear constant head boundary:
27fTH
o
Qo = l n ( ~ )
with Qo discharge rate along the investigated section (m
3
/s), T hydraulic
transmissivity (m
2
/s), Ho piezometric level above gallery (m), r radius of gallery
(m). Hydraulic conductivity k is then estimated from T = k x e, with e width of
investigated zone (m). A range of potential hydraulic conductivities was calcu-
lated for fixed heads at the topographic surface (z) and at a piezometric level of
0.7z. The latter assumption is based on observations in similar rock types and
overburden in the scope of the Gotthard-basetunnel project of AlpTransit/NEAT
in Switzerland (Klla 1993). Table 3.3 summarizes the estimated values.
In the scope of the AlpTransit/NEAT project (Loewet al. 2000), a compre-
hensive database of effective hydraulic parameters from the geological units of the
Aar and Gotthard Massif has been compiled (Colenco 1993). The derivation of
effective transmissivities was thereby based on the observation of discharge rates
to numerous existing tunnels and subsurface galleries in these formations (Loew
et al. 1996). Based on these, effective hydraulic conductivities were calculated far
Ionger tunnel sections (>100 m, Voborny et al. (1994)). The so derived hydraulic
conductivity for Hercynian granites in the Gotthard Massif such as the Rotondo
Granite at the tunnel elevation is 1.1 x 10-
8
m/s. As no direct observational data
was available in the scope of this study to constrain the hydraulic conductivities
of the neighboring geological units to the Rotondo granite, the above database
(Colenco 1993) was used to establish initial approximations for the numerical
model (Table 3.4).
Several studies have shown the depth-dependency of the hydraulic conduc-
tivity/transmissivity in crystalline rocks (Snow 1968; Herbert et al. 1991; Thury
et al. 1994). Furthermore it can be observed that a distinct hydraulic discontinu-
ity occurs in the uppermost 100-200 m. Close to the terrain surface substantially
higher conductivities due to denser fracturation and higher apertures occur. This
can be explained by stress release, weathering processes and gravitationally in-
duced weakening of the bedrock along steep slopes. This zone if often referred to
as the decompressed zone (Jamier 1975; Raven 1977; Cruchet 1985) and can also
48
be observed in the research area (Ofterdinger 2001). On the basis of a review
of studies in the crystalline rocks of the Swiss Alps and the crystalline basement
of Switzerland, Marechal (1998) proposes an exponential decrease of hydraulic
conductivity, where k (mjs) at a given overburden C (m) follows:
k = k
o
x e-
axC
(3.2)
with k
o
being the hydraulic conductivity at surface. Marechal (1998) derives
a value of 0.05 rn-I for the exponent a for the uppermost decompressed zone
and a value of 0.005 rn-I at depth. This approach has been adopted in this
study, defining the decompressed zone as the uppermost 100 m of the massif.
Although the depth dependency is frequently observed for the gneiss-formations
within the Aar- and Gotthard-massif, this clear dependency is not observed within
the Hercynian granites such as the Rotondo, Fibbia and CacciolajGamsboden
granite (Colenco 1993; Loew et al. 1996). Furthermore it can be observed that
major water conducting faults often also lack this depth-dependency (Thury et al.
1994). Therefore, the hydraulic conductivity far the Hercynian granites was not
modified with depth, except the distinction between the decompressed zone and
the deeper subsurface.
The preferred orientation of fractures and faults, associated with inflows to
the tunnel within the Rotondo granite and the pronounced structural anisotropy
within the neighboring gneisses (foliation) imparts an anisotropy on the hydraulic
conductivity of these formations, promoting flow along strike of these features.
As their orientation is approximately sub-parallel or at acute angles to the over-
all strike of the Gotthard Massif, the anisotropy is simplified to be orientated
globally with k
l
> k
2
, whereby k
l
is orientated parallel to the overall strike ofthe
massif (72
0
). As the structural features, such as the observed water conducting
fractures within the granite as well as the foliation in the gneisses are steeply
to sub-vertically inclined, the vertical component of hydraulic conductivity was
set equal to the maximum horizontal component. Studies in similar rock-types
suggest anisotropy ratios in the range of 1:::;10-170 (Hsieh et al. 1985; Raven
1985; Nagra 1988). In the absence of furt her constrains on the absolute value of
the anisotropy ratio a conservative approximation of a 1:10 ratio was taken for
the numerical model.
49
Table 3.3 shows that the analytical approximation of the k-value for the fault
zone is ",1 order of magnitude larger than the k-value approximated for the re-
maining tunnel section. Other studies have stated permeability contrasts between
fault zone and neighboring bedrock to be in the range of up to 3 orders of mag-
nitude (Smith et al. 1990; Forster et al. 1994; Caine et al. 1996)). However,
most of these studies consider narrow discrete fault zones composed of the suc-
cession of fault gouge, damage zone and undamaged protolith. As stated earlier,
in this study the term fault zone is attributed to a 120 m wide zone comprising
several narrow discrete faults with sections of densely fractured rock in between.
Because of this, the contrast to the surrounding bedrock is expected to be less
pronounced compared to studies, where a specific element of a discrete fault zone
in the strict sense, e.g. the damage zone is compared with the essentially unfrac-
tured bedrock. Therefore a conductivity contrast of 1:10 is initially set between
the Rotondo granite as such and the fault zone within the granite. A similar con-
trast has been found by Colenco (1993) and has been applied in another study
by Herbert et al. (1991) for a regional model.
Anisotropy within the fault zone is set to be k
1
> k
2
, whereby k
1
is aligned
parallel to the strike of the fault zone. Similar to the granite bedrock, the vertical
component of hydraulic conductivity (k
3
) was set to k
3
= k
1
, as the fault zone is
subvertical inclined. Many studies have shown the decreased fault-normal flow
across fault zones, mainly caused by the low permeable central fault gouge and
also by the preferred fault-parallel orientation of fracturation within the adjacent
damage zone (Caine et al. 1996; Smith et al. 1990; Lopez and Smith 1996). This
anisotropy ratio has been shown to range between ::;10 (Forster and Evans 1991;
Andersson et al. 1991) up to several orders of magnitude (Evans et al. 1997;
Smith et al. 1990). As the fault zone in our case includes only very few gouge
material, which would potentially inhibit fault-normal flow, an anisotropy ratio
within the fault zone of 1:10 was initially assumed.
3.3.4 Ground water recharge rates
Depending on topography, landuse, vegetation and slope exposition, ground water
recharge shows strong variability across catchments (e.g. Lerner et al. 1990). As
the research area is situated in a high mountainous area covering a wide range of
50
elevation zones, the spatially discretization of the ground water recharge rates is
a crucial factor for estimating the upper boundary condition of the ground water
flow model. For this reason, the recharge rates applied in the ground water flow
model were extracted from the results of a calibrated hydrological model of the
study catchment. The applied GIS-based model PREVAH (precipitation-runoff-
evapotranspiration-hydrotope model) combines the spatially differentiation of hy-
drologically similar response units (HRU) or hydrotopes and a runoff generation
concept that allows the separate ca1culation of the water balance within each
hydrotope (Gurtz et al. 1999). The hydrotopes were defined to represent hydro-
logically homogeneous areas according to the most important factors controlling
evapotranspiration and runoff formation process, such as the meteorological in-
puts, topography (catchment area, altitude, exposition and slope), landuse and
soil characteristics (Gurtz et al. 1990). For the glaciated areas a conceptual
model was applied which was developed for a research study in the Rhne area
(Badoux 1999). This concept attributes meltwater from the glaciated areas below
the equilibrium line (ELA) preferentially to the fast runoff storage, while melt-
water originating above the ELA preferentially contributes to the slow runoff
storage, i.e. to ground water recharge.
The hydrological model was first calibrated and validated in the area of
main interest, the Gerental basin (rv 40 km
2
), and then applied to the larger
ground water flow model domain using identicalmodel parameters as calibrated
in the Gerental basin. The hydrograph simulation for the gauging station at the
Gerental catchment outlet were performed for the period January 1991 to October
1999. The results of the hydrograph simulation are discussed elsewhere (Vitvar
and Gurtz 1999; Vitvar and Ofterdinger 2001). The temporal variations of the
simulated recharge rates show maximum recharge rates in the upper Gerental-
valley during spring and summer (April-September), commencing with the onset
of the melting period. The resulting recharge distribution shows a strong spatial
variability, especially in the high altitude regions of steep bare rock slopes, where
low recharge areas contrast to the upper regions of the glaciers, characterized by
high recharge rates within small restricted areas (Figure 3.5 and 3.6). Another
striking feature in the recharge distribution is the observation of low to moderate
recharge rates along the valley floor and the lower valley slopes. From a hydro-
dynamic perspective these areas are expected to constitute potential exfiltration
51
areas, where essentially no ground water recharge should occur. This conceptual
discrepancy might lead to a systematic error in the spatial distribution of the
recharge rates, as with regard to the overall water budget, the wrongly assigned
infiltrating waters in the potential exfiltration zones would need to be attributed
to other parts of the catchment to achieve a similar hydrograph simulation. For
the purpose of the numerical simulations an average recharge rate distribution
was extracted from the 9-year average. The spatial average recharge rate of this
distribution for the whole model domain is 12.05x 10-
4
m/d, which is in agree-
ment with studies in similar topographie and geologieal settings in the Central
Alps (Klla 1993) giving a range of2.7xl0-
C
21.9xl0-
4
m/d with a most prob-
able value of 10.9x 10-
4
m/d.
3.3.5 Flow porosities
For the ca1culation of isochrones in the subsequent modeling an estimate of the
flow porosity (kinematic porosity) nf for the fractured Rotondo granite was re-
quired. This flow porosity was estimated from the transmissivity data obtained
in the borehole tests and the frequencies of the fracture sets mainly associated to
tunnel inflows. First the apparent hydraulic aperture e of the flowing fractures
intersected by the borehole was estimated based on the cubic law:
e = ~ 1 2 T I ' (3.3)
pg
with T hydraulic transmissivity (m
2
/s), Jl fluid viscosity (kg/ms), p (kg/m
3
)
fluid density and g (m/s
2
) gravitational constant. The results are then used to
estimate a range of flow porosities according to:
e
nf = - (3.4)
L
where L is the measured average flowing fracture spacing (Himmelsbach et al.
1998). Comparison ofthese results were also made to fracture porosities approx-
imated from the empirical relation of Snow (1968):
(3.5)
52
with k
i
intrinsic permeability (m
2
) and L average flowing fracture spacing
(m). As L was set to equal the spacing of the fracture sets observed to be mainly
associated to tunnel inflows and no discrimination within these sets of the spacing
of specific flowing fractures was done, the flow porosities derived on the basis of
L represent a potential overestimation of the actual flow porosity.
The so ca1culated values are assumed to represent the flow porosities of the
more intact granite sections with moderate fracturation (blocks,nfB ). Ca1culated
values for nfB range from 2.4x 10-
5
to 2.9x 10-
4
and are in agreement with
published values for similar rock types (Abelin et al. 1991; Guimera and Carrera
1997).
For the approximation of an average flow porosity for the Rotondo granite on
a larger scale, the discrete smaIl-scale faults as weIl as the studied major fault
zone (which in turn also consists of several discrete small scale faults) within the
granite also have to be accounted for. Flow porosities for these features (nfF)
could not be directly estimated but published values for nfF indicate a range
from 1.3x 10-
3
(Himmelsbach et al. 1998) to 7.4x 10-
3
(Frick 1994).
Based on these values, an average flow porosity for the Rotondo granite on
the regional scale (nf RG) was approximated by averaging the flow porosities of
the faults and of the granite blocks along the rv1.5 km tunnel profile, weighted
by the measured width of these sections along the tunnel profile following:
L:W
F
x nfF + L:W
B
x nfB (3.6)
nfRG = W
T
where W
F
(m) is the measured width of the encountered smaIl-scale faults
along the tunnel section and WB (m) the width of the intermediate more intact
blocks of granite. W
T
(m) is the total length of the tunnel profile. Inserting the
above range of values for nfB and nfF yields an estimate for the average flow
porosity of the Rotondo granite nfRG ranging from 2.0x10-
4
to 1.0x10-
3
3.4 Conceptual model

Many approaches are currently available for the simulation of flow in fractured
rocks (NRC 1996; Smith et al. 1990), each corresponding to a different concep-
tualization of the medium, ranging from viewing the medium as an equivalent
continuum with either single or double porosity based on prescribing effective
53
hydraulic parameters (e.g. Carrera et al. (1990)) to associating flow only to
fractures with the need for defining the geometry of the fracture network (Long
et al. 1982; Cacas et al. 1990; Dverstorp and Andersson 1989). A combination of
the two above approaches, i.e. so-called hybrid models, consists of treating minar
fractures and the rock matrix as a continuum and explicitly modeling major frac-
ture zones (Kimmermeier et al. 1985; Carrera and Heredia 1987). In addition the
channeling network concept reduces the 2d flux through the fracture planes to a
1d channellying in the fracture plane (Neretnieks 1983). As this study is focused
on the ground water flow systems on the scale of the entire Rotondo massif and
its interactions with the subsurface galleries we choose a continuum approach
with one discrete model fault zone to analyze the ground water flow at this scale.
Effective hydraulic conductivities are introduced into the model to characterize
the hydrogeological units. A fully three-dimensional model is chosen to simulate
the influence of the complex topography and the subsurface galleries on the flow
systems. The effect of spatially distributed ground water recharge rates is studied
by assigning a free and movable water table. The investigated model fault zone
within the Rotondo granite is represented in a discrete manner with three dimen-
sional elements having anisotropic hydraulic parameters. Hydraulic anisotropy
of the bedrock is introduced based on field observations of structural anisotropy
and preferred orientations of water conducting features.
The focus of the investigation and thus the modeling effort is aimed at the
investigation of the ground water flow systems within the Rotondo granite. How-
ever, as the position ofthe water divides is uncertain due to the anticipated effects
of the tunnel galleries and the complex topography, the model domain is later-
ally extended to include neighboring formations and topography. Even though
this introduces additional uncertainty into the model due to badly constrained
effective hydraulic parameters for these units, we consider this to be less crucial
than prescribing wrong boundary conditions close to the area of interest.
3.5 Numerical Model
The simulations of the ground water flow in the research area were carried out
with the finite element FEFLOW-code (Diersch 1998). The mesh consists of
54
141'022 nodes and 259'156 elements distributed over 13 layers (Figure 3.7 and
3.8). Topography was extracted from the digital elevation model on a 25x25
m grid. The model base is at 0 m a.s.l. The total projected area of the model
domain comprises ",262 km
2
. The mesh was densified along the tunnel traces, in
areas of complex topography and along the trace of the model fault zone and the
rivers/streams. The top three layers ofthe model were kept sub-parallel to topog-
raphy, while the layers below are referenced in orientation sub-parallel to the trace
of the subsurface galleries (Figure 3.7). This vertical discretization was chosen
to enable a better resolution of the free & movable water table, to assign depth
dependent k-values and to achieve a densified discretization around the galleries.
The outer boundary conditions of the domain are set impervious (no-flow condi-
tions). The perennial streams and rivers are treated as fixed head inner boundary
conditions. From the results of the hydrological model the spatially distributed
long-term mean recharge rates were extracted as upper boundary condition for
the model. The Furka-basetunnel and the Bedretto-tunnel are represented as line
sinks with fixed elevation heads. Initial conditions for steady-state simulations
are fully saturated conditions.
3.5.1 Modeling approach
The stable behaviour over time of the parameters measured in the subsurface
(inflow rates and hydrochemical/isotopic composition of ground water (Ofter-
dinger et al. 2000) indicate that steady-state simulations are a feasible approach
to calibrate effective k-values. Calibration efforts are carried out in a trial and
error procedure and are focused on the Rotondo granite and the immediate ad-
jacent formations in order to reach a correlation to the observed inflow rate to
the Bedretto-Tunnel and the Furka-basetunnel. In a first phase, k-values for the
Rotondo granite and/or the fault zone within the granite are adopted to achieve
the observed discharge rates to the Bedretto-tunnel. Secondly, to match the
measured inflows to the Furka-Basetunnel, especially along the section through
the northern margin of the granite body, additionally k-values for the northern
gneisses (0, PN) are adjusted to achieve a model calibration. The recorded head
data from the research borehole and the position of the spring line on the ter-
rain surface serve as additional criteria to assess the plausibility of the modeling
results. During the modeling process three base case scenarios are evaluated in
55
which we address the impact of the model fault zone and the subsurface galleries
on the ground water flow system.
In scenario 1, a calibration is sought for a continuum numerical model without
inc1uding the higher conductive fault zone.
In scenario 2, an isotropic higher conductive fault zone (FZ) is incorporated
into the model with a k-value contrast to the surrounding granite bedrock (G) of
1:10 (kFZ = 10 X k
lG
).
In scenario 3, anisotropy is introduced within the fault zone of scenario 2.
Anisotropy ratio is set to k
1FZ
= k
3FZ
= 10 X k
2FZ
, with k
2
representing fault-
normal flow direction.
The influence of the subsurface galleries on the ground water flow field within
the Rotondo Granite is investigated for all model scenarios by removing the fixed
head inner boundary conditions associated with the galleries.
In a sensitivity study following the model calibration of the individual scenar-
ios the influence of varying recharge conditions is addressed. Model simulations
with aseries of spatially uniform recharge rates are compared to the model re-
sults with spatially distributed ground water recharge rates. As values for the
uniform recharge rates the spatial average recharge rate for the whole model
domain (12.05x 10-
4
m/d or 19 %of mean annual precipitation), the spatial av-
erage recharge rate for the upper Gerental catchment (region above the Bedretto-
tunnel, 5.75x10-
4
m/d or 7 %ofmean annual precipitation) as weIl as a minimal
value of2.74x10-
4
(
rv
3 %ofmean annual precipitation) were chosen. These val-
ues lie in the range of published recharge rates for granitic terrain, ranging from
2-21 %of mean annual precipitation (Lerner et al. 1990; KIla 1993).
For all model variants partic1e tracking is carried out to delineate the respec-
tive recharge areas to the Bedretto-tunnel. Partic1e backtracking is started along
the nodes representing the Bedretto-tunnel (on slice 8, Figure 3.7) as weIl as
along the projected nodes on the 5 m over- and underlying slice. Additionally,
56
forward tracking was initiated from the top slice of the model to cross-check the
delineated recharge areas. Finally, isochrones were ca1culated along the pathlines,
initiated at the tunnel nodes, to approximate the advective travel times across
delineated recharge areas to the Bedretto-tunnel.
3.5.2 Model calibration
3.5.2.1 Scenario 1
To reach the calibration target of ground water infiow to the Bedretto-tunnel,
the k-value of the Rotondo granite had to be adjusted to k
1
=1.2x10-
8
m/s
(Table 3.5). This value may range from 1.15xlO-
8
to 1.25x10-
8
m/s to still
meet the calibration target. To achieve a calibration for the granite section
along the Furka-basetunnel, the k-value of the northern adjacent gneiss-unit (0)
was additionally reduced by a factor of 0.4 and the k-value of the neighboring
PN-unit had to be reduced by one order of magnitude from the initial value to
match the fiow rates at the Oberwald-Portal of the Furka-basetunnel. To furt her
match the infiows along the section of the Furka-basetunnel which drains towards
the Realp-Portal (Figure 3.4), the k-value for the Cacciola-granite was increased
to k
1
=8.0 X 10-
8
m/s. This is regarded as plausible since major infiows along
the Cacciola section of the Furka-basetunnel are reported (Keller and Schneider
1982). The calibrated k-value for the Rotondo granite is in good agreement with
the first approximations following the analytical solution (Table 3.3). Simulated
heads in the monitoring intervals are underestimated by the fiux-calibrated model
(Table 3.5). Furthermore, Figure 3.9 shows that this model scenario without the
fault zone is not able to reproduce the cumulative infiows along the Bedretto-
tunnel.
3.5.2.2 Scenario 2
With the fixed k-value contrast between the granite bedrock and the fault zone of
1:10, a calibration to the infiow rate along the fault zone section of the Bedretto-
tunnel as well as to the infiow rate along the remaining bedrock section was
reached with k-values of 6.0x10-
9
and 6.0x10-
8
m/s for k
1
of the granite and
the isotropie fault zone respectively. These values are again in good agreement
with the analytical approximations. Variation of the respective k-value for the
57
granite bedrock and the fault zone within one order of magnitude yield inflow
rates to the Bedretto-tunnel beyond the calibration target. An increase of the
absolute k-values of the granite and the fault zone to 8.0x 10-
9
18.0x 10-
8
m/s
leads for example to inflow rates to the Bedretto-tunnel and the fault zone section
of 19.3 and 10.31/s respectively. A decrease ofk-values to 5.0xlO-
9
/5.0xlO-
8
yields discharge rates of 12.3 and 6.7 1/s for total inflow to the Bedretto-tunnel
and the fault zone section. To achieve a match to the inflows along the Rotondo
granite section of the Furka-basetunnel, the k-values of the neighboring gneiss
(0) had to be increased compared to scenario 1 to a factor of 0.6 from the initial
value. The cumulative inflow along the Bedretto-tunnel for the calibrated model
approximates weIl the measured inflow profile (Figure 3.9). However, similar to
scenario 1, the heads in the monitoring intervals are underestimated (Table 3.5).
3.5.2.3 Scenario 3
The introduction of anisotropy within the fault zone (base case k
1
= 10k
2
with
k
2
fault-normal orientated) into the calibrated model of scenario 2 causes only
a slight decrease in fluxes to the Bedretto-tunnel (-0.8 l/s for the total tunnel
from which -0.6 1/s along the fault zone section). K-values were adjusted to
6.5x 10-
9
and 6.5x 10-
8
m/s for the Rotondo granite and the fault zone respec-
tively to reach the calibration targets. A small additional increase of the k-values
to 7.0x10-
9
/7.0x10-
8
m/s leads to tunnel inflows beyond the calibration target
of 16.3 and 8.4 l/s for the Bedretto-tunnel and the fault zone section respectively.
The cumulative inflow along the Bedretto-tunnel resembles the simulation results
of scenario 2. Like in the previous scenarios the model fails to reproduce the local
head measurements in the monitoring intervals (Table 3.5).
3.5.3 Model results
In all model scenarios a considerable decline of the water table below topography
is simulated along the crests within the Rotondo area, especially along the trace of
the subsurface galleries. For scenario 1 simulated position of the water table range
from ::;200 m below the ridges above the Furka-basetunnel in the orthogneiss to
up to 600 m below the crests of the Saashrner ridge in the Rotondo granite
(Figure 3.10). In the east of the model domain deep ground water tables can
58
be observed in the Winterhorn and Siroerbenhorn area. In the western part of
the domain the ground water table lies at 100-300 m below the ridges of the
Mettligrat and Mittaghorn. In the Blauberg area to the north of the domain,
depth to the ground water table reaches 200 m.
In scenario 2, the dec1ine of the water table below topography is less pro-
nounced. Depth to the ground water table reaches up to 300 m below the
Saashrner and up to 250 m above the Furka-basetunnel in the Rotlligrat area
(Figure 3.11). No additional dec1ine of the water table due to the introduction
of the higher conductive fault zone can be observed. The position of the ground
water table for scenario 3 resembles c10sely the simulation results of scenario 2.
Even though the water table position in the core of the massifs differ signif-
icantly between model scenarios the observed altitude of the spring discharges
in the study area (Figure 3.2) is reproduced by all model scenarios within an
accuracy of 20-100 m. The inaccuracy leads thereby to an overestimation of the
spring line altitude.
For all three model scenarios, the partic1e tracking delineates recharge areas
to the Bedretto-tunnellying mainly along the crests of the high mountain ridges.
The upper glacier regions present here essentially the main recharge source to
the gallery (Figure 3.12). An additional recharge area is situated along the mid-
altitude southern slope of the Sasshrner, along a smaller valley inward trending
ridge. Here moderate recharge rates are predicted by the hydrological model
(Figure 3.6). In all three scenarios, the northern end of the Bedretto-tunnel is
recharged from the glaciated areas of the Leckihorn and Saashrner area as weIl
as from the above mentioned recharge area along the Saashrner slope. The fault
zone section along the Bedretto-tunnel is recharged from the glaciated area at
the Witenwasserenpass area (between Leckihorn and Witenwasserenstock) in the
east and from the glaciated regions on the north-facing slope of the Rotondo.
The introduction of the higher conductive fault zone in scenarios 2 and 3 into
the model leads to an enlargement of the recharge area, in particular in the upper
regions of the Witenwasserenglacier. Flow entering the fault zone is aligned with
the fault zone orientation and directed to its intersection with the gallery. For the
southern end of the accessible part of the Bedretto-tunnel, the particle tracking
indicate the upper region of the Gerenglacier as recharge area.
59
To assess the influence of the subsurface galleries on the ground water flow
system in the Rotondo area all calibrated model scenarios were additionally sim-
ulated without the fixed head boundary condition along the tunnel profile. Sim-
ulated heads without the tunnel condition show significantly less dec1ine of the
water table below ground surface. For scenario 1 this head difference between
simulations with and without tunnel amounts to 350 m in the Saashrner area,
50-150 m along the Rotlligrat and rv50 m in the Siroerbenhorn area (Figure
3.15). For scenario 2 and 3 theses differences range from up to 150 m along the
Saashrner to 50-100 m and 50-150 m along the Rotlligrat and Siroerbenhorn
respectively. With regard to the partic1e tracking all model scenarios show that
without the tunnel gallery, only the eastern part of the above described recharge
areas, i.e. the Leckihorn- to Gerenglacier-area, is drained towards the tunnel
position. This influence of the gallery on the natural flow field is also illustrated
in Figure 3.16, where the influence of the Bedretto-tunnel is shown to cause a
strong modification of head gradients.
Advective travel times from the individual recharge areas were estimated from
the calculated isochrones along the pathlines, initiated at the tunnel nodes, and
are given in Table 3.6. The range of values for the applied flow porosities nfHG
yields a wide range of 1 to 14 years for flow from individual recharge areas to the
gallery (Table 3.6). In a relative comparison amongst the estimated travel times
the rapid flow path via the fault zone is apparent (Figure 3.13 and 3.14).
3.5.4 Sensitivity study
The partic1e tracking for the individual model scenarios show that the prescribed
distributed recharge rates strongly influence the position of the tunnel recharge
areas and type of recharge source, i.e. recharge from glaciers or mountain slopes.
Recharge is hereby dominated by the high recharge rates in the upper regions of
the glaciers. To assess the impact of a change in the recharge conditions, three
model variants with different uniformly distributed recharge rates are considered
and applied exemplarily to model scenario 2. As described before, recharge rates
are thereby varied within a plausible range of potential recharge rates. Note that
simulations with uniform recharge rates are referred to in the following as model
60
variant opposed to the calibrated simulations (model scenarios).
First we consider a uniform recharge rate of 5.75x 10-
4
m/d, a value extracted
from the distributed recharge rates as mean recharge rate in the upper Gerental
catchment above the Bedretto-tunnel (>2000 m a.s.l., variant 1). Figure 3.17
shows in comparison to Figure 3.11 that the position of the water table is signifi-
cantly modified. No dec1ine of the water table is observed in the Saashrner area
while an additional dec1ine of the water table is located in the Witenwasseren-
stock and Rotondo area. An additional dec1ine of the water table can furthermore
be observed along the ridges in the east and west of the domain as weIl as in
the Blauberg area to the north. This shows the strong effect of the spatially dis-
tributed recharge rates on the position of the ground water table especially in the
Saashrner area. Ground water fluxes to the Bedretto-tunnel are only changed
by 2-3 %and fluxes to the Rotondo granite section of the Furka-basetunnel by
7% (Table 3.7). Changes in the simulated heads in the monitoring intervals are
only minor (Table 3.7). The position of the recharge areas to the Bedretto-tunnel
is strongly affected by the change in recharge conditions, especially to the west of
the gallery (Figure 3.12). Here, the recharge area is now situated along the entire
ridge of the Saashrner. No more drainage from the Saas- and Muttenglacier
occurs. The recharge area to the east of the Bedretto-tunnel is shifted downslope
in the Leckihorn area towards the west. Recharge areas to the fault zone are still
situated along the glaciated area of the Witenwasserenpass, with an increased
area towards lower elevations, and along the southern Rotondo slope.
With a higher uniform recharge rate of 12.05 x 10-
4
m/d (variant 2), extraeted
from the distributed recharge rates as mean recharge rate for the entire model
domain, generally less dec1ine of the water table below the ground surface can be
observed when compared with the results from model scenario 2. No dec1ine of
the water table below the Saashrner ridge can be observed. Also the depth to
the ground water table in the Rotlligrat area is decreased. In the eastern and
western parts of the model domain (Winterhorn and Mettligrat area) changes
are only minor. Changes in the ground water flux to the galleries are nearly in-
significant with 1-2 % in the Bedretto-tunnel and 1-5 % in the Furka-basetunnel
(Table 3.7). Changes in simulated heads for the monitoring intervals are also
61
only minor. The delineated recharge areas to the Bedretto-tunnel are similar to
those deduced for the model variant 1.
In a third variant, the uniform ground water recharge rates are reduced to
2.74 X10-
4
m/d, a value at the lower end of plausible recharge rates. This low
uniform recharge rate has a dramatic effect on the position of the ground water
table elevation in the model domain (Figure 3.18). Large areas, not only restrieted
to the crest regions of the mountain ridges, show considerable decline of the water
table below the ground surface. The water table is in fact declined closely to
the valley bottom in the upper Gerental catchment with considerable decline
also along the trace of the Bedretto-tunnel across the upper Gerental valley.
The depth to the ground water table amounts to 400-600 m in the Winterhorn,
Saashrner and Siroerbenhorn area. In the Leckihorn and Witenwasserenstock
area, depth to the water table also reach 400-600 m. The decline of the water
table extends below the Mutten- and Witenwasserenglacier and reaches 100-200
m above the Bedretto-tunnel. The effect of this change in recharge rate is also
manifested in a clear change in ground water fluxes to the Bedretto and Furka-
basetunnel with changes of 9-10 % and 12-30 % respectively (Table 3.7). Also
the simulated heads in the monitoring intervals show a significant decrease (Table
3.7). The recharge areas are drastically modified. To the west, a large area along
the Saashrner slope is recharging to the Bedretto-tunnel. Likewise to the east,
the slopes of the Leckihorn are recharging to the gallery. Essentially a broad area
from directly above the tunnel trace and expanding to the west to the Sasshrner
and to the east to the Leckihorn area are drained to the Bedretto-tunnel. The
recharge area to the fault zone is still largely situated along the Witenwasseren-
and Gerenglacier.
3.5.5 Discussion of model results
The modeling of the above scenarios in steady-state simulations show that the
calibrated k-values are generally in good agreement with the analytical apriori
estimates. With regard to the hydraulic conductivities of the individual geological
units, the modeling process showed that the fluxes to the Bedretto-tunnel are
mainly governed by the prescribed k-value within the Rotondo granite, while
the inflows along the granite seetion to the Furka-basetunnel are additionally
62
sensitive to the k-values of the orthogneisses. Changes in hydraulic conductivity
of the Rotondo granite in the range of less than one order of magnitude cause
hereby observable differences in the elevation of the ground water table, especially
in the Saashrner area.
Figure 3.19 illustrates regional and local ground water flow systems within
the Rotondo area whereby the regional scale ground water flow is mainly directed
towards the Rhne-valley in the east and the Ticino-valley at the southeastern
border of the model domain.
Unsaturated zones below the ridges and crests of up to several hundred meters
are observed (Figure 3.10 and 3.11). The depth to the ground water table below
the Saashrner as weIl as along the eastward section of the Furka-basetunnel are
enhanced by the presence of the subsurface galleries (Figure 3.15). Simulated
pronounced dec1ine of the water table in the Saashrner area are confirmed by
the findings of a Nuc1ear-Magnetic-Resonance-survey (NMR) along the southern
slopes of this ridge, indicating a depth to the water table of >90 m (Legchenko
et al. 1998). The presence of the subsurface galleries in the Rotondo area causes
strong modifications of the natural head gradients, leading locally to areversal
of natural ground water flow directions (Figure 3.16).
The introduction of a higher conductive fault zone within the base case sce-
nario leads to a modification of the head distribution in the model domain, as
the hydraulic conductivity of the surrounding bedrock was adopted to match the
measured flow rates to the galleries. Figure 3.20 and 3.21 indicate how the in-
troduction of the fault zone leads to a modification of lateral flow components
and the position of local water devides. The ratio of anisotropy within the fault
zone is of only little significance in the simulated scenarios. This is most prob-
ably linked to the overall orientation of the fault zone sub-parallel to the main
flow field and the regional anisotropy of hydraulic conductivity within the adja-
cent bedrock. Thus a potential barrier effect in fault-normal direction due to a
high ratio of anisotropy within the fault zone would not significantly hinder the
natural fault-parallel oriented flow field.
Heads in the monitoring intervals of the research borehole, characterizing the
head distribution in a moderately fractured part of the granite body, are un-
derestimated in all flux-calibrated model scenarios. Modifications of the k-value
contrast between the granite bedrock and the fault zone from the originally set
63
contrast of 1:10 does not provide a better match ofboth fiuxes and the local head
data. For example, a decrease of the k-value of the Rotondo granite in scenario
2 to a value of 1.0x 10-
9
m/s while maintaining the hydraulic conductivity of
the fault zone at 6.0x 10-
8
leads only to an increase of simulated heads in the
monitoring intervals of 21 m (interval 1) and 17m (interval 2) respectively. At
the same time, simulated ground water infiow to the Bedretto-tunnel is already
reduced by as much as 5.4l/s. This amount can not be compensated for by a
potential increase of the fault zone k-value without exceeding the measured cali-
bration target for the infiow along the fault zone section. This model behaviour
indicates that the presented equivalent continuum approach with only a single
discrete higher conductive fault zone is not able to simultaneously reproduce the
measured fiuxes to the gallery and the local head data from the research bore-
hole correct1y. For a match of both fiux and head data, further discrete higher
conductive features would need to be inc1uded into the model. This would then
essentially lead to lower k-values for the granite bedrock in the calibration process
and thus to an increase of simulated heads at the borehole site.
Another explanation for the discrepancy between measured and simulated
heads in the monitoring intervals might also be that the natural ground water
fiow systems have not yet reached a steady-state as assumed for the model simula-
tions. The validity of the assumption of steady state conditions for the numerical
modeling was tested applying the analytical solution of a constant head test in an
ideal (homogeneous, isotropic, etc.) confined aquifer with a weIl of finite radius
(r
w
) and a constant head boundary at a distance r (see Appendix F). In our case
the tunnel represents a constant head sink with radius r w while a constant head
boundary is imposed at the topography at distance r to the tunnel. Following
the approach of Murdoch and Franco (1994) yields the dimensional fiow rates
qd over dimensionless time td as depicted in Figure 3.22 for varying dimension-
less distances to the constant head boundary h
d
= r / rw' Also depicted are the
asymptotic values ql representing steady state conditions for the corresponding
solutions. Figure 3.23 illustrates the residual percentage between ql and qd over td
for our case with a value of h
d
= 800. From this it can be seen that a steady state
is reached at approximately td = 1 X 10
8
. Solving the relationship for td utilizing
the values of transmissivity deducted from the numerical model and proposing a
range of storativities from 10
6
to 10
8
yields values of t
d
after aperiod of t = 20
64
years, time since tunnel construction, in the order of 2.4 x 10
9
to 2.4 X 10
11
. This
illustrates that the application of steady state conditions for the modeling is a
valid assumption.
Generally, the model simulations reproduce weIl the position of spring dis-
charge in the upper Gerental vaIley. As these are situated only a few decameters
to hundred meters above the valley floor along the steep slopes, the influence
of the conceptual discrepancy between the modeled exfiltration zones and the
recharge areas predicted along the valley floor by the hydrological model as stated
earlier is thought to be less significant. The area concerned is rather small com-
pared to the scale of the massif.
Applying the ground water recharge rate distribution from the hydrological
model to the individual model scenarios yields similar catchment areas to the
Bedretto-tunnel. In the framework of this prescribed recharge distribution these
are mainly associated to the glaciated areas with high recharge rates. The inflows
to the fault zone section along the Bedretto-tunnel are in all model scenarios con-
sistently recharged by the glaciated areas of the Witenwasser- and Gerenglacier.
The distribution of delineated recharge areas is shown to be strongly depen-
dent on the spatial distribution of the recharge rates within the model domain.
In the prescribed distributed recharge rates the most striking feature is the strong
contrast at high altitudes between the steep bare rock slopes with little or essen-
tially zero recharge and the upper regions of the glaciers, where high recharge
rates are predicted (Figure 3.5and 3.6). These strongly localized high recharge
rates dominate the delineated recharge areas.
The effect of the spatial discretization of the recharge rates is apparent when
compared to the results from the simulations with an equivalent uniform mean
recharge rate. The recharge areas to the Bedretto-tunnel are now more broadly
aligned along the mountain ridges and especially to the west a wider area along
the Saashrner ridge is recharging to the Bedretto-tunnel (Figure 3.12). With a
uniform recharge rate the fault zone section in the tunnel is consistently recharged
from the area of the Witenwasseren- and Gerenglacier. Additionally, the lower
slopes of the Gerenglacier are recharged along the fault zone to the Bedretto-
tunnel. The ground water fluxes to the Bedretto-tunnel are only affected to a
minor degree, when an average uniform recharge rate is extracted from the spa-
tially distributed recharge rates of the upper Gerental (variant 1) or the whole
65
model domain (variant 2) respectively. However, when the absolute value of the
uniform recharge rate is lowered to a minimum plausible value (variant 3), dra-
matic changes in terms of recharge area, water table position and tunnel inflows
are observed (Figure 3.18, Table 3.7). The latter are decreased c1early below
measured inflow rates to the galleries. Forster and Smith (1988a) investigate
the non-linear fashion in which the water table elevations respond to changes
in ground water recharge rate. The response of the water table elevation in the
Rotondo granite (below the Saashrner) appears to be illustrating this non-linear
behaviour.
Comparing the position of the water table for the individual model scenarios
and variants (Figure 3.10, 3.11, 3.17 and 3.18), the magnitude of variations in
water table elevation below the Saashrner is striking. One explanation for this
sensitive behaviour lies in the specific topography, i.e. the high relief of this
mountain. The Saashrner are bound to three sides by deep valleys, especially
to the north and west with steep slopes ranging from 1800 to 2900 m.a.sI. To the
east the subsurface galleries impose an additionallow fixed head boundary. The
surrounding crests like the Leckihorn and Rotlligrat are aligned along elongated
mountain ridges, show less steep relief and bounding valleys are generally at an
altitude >2000-2100 m a.s.I. Model simulations of Forster and Smith (1988a)
have also shown that high-relief terrain amplifies the impact of factors such as
permeability and recharge rate on ground water flow systems within mountain
massifs.
As a consequence of the sensitivity study, the eastern recharge areas can be
regarded as robust, since they are delineated by all model variants as recharge
areas, independent of the recharge conditions. The position of the recharge ar-
eas to the west of the Bedretto-tunnel are more c1early subject to the spatial
discretization of the recharge rates and some ambiguity remains on their posi-
tion. In the following, we will discuss how the environmental isotope data can
supplement these interpretations.
66
3.6 Comparison to results from environmental
isotope study
In the previous study of the isotopic and hydrochemical composition of the en-
countered ground water along the Bedretto-tunnel (Ofterdinger et al. 2000),
measured 6"
18
0 values indicate an altitude effect with higher recharge altitudes
towards the northern end of the Bedretto-tunnel (Figure 3.24). Estimated mean
recharge altitudes lie in the range of 2600-2800 m a.s.l. The numerical simula-
tions with distributed recharge rates yield recharge areas to the northern end of
the Bedretto-tunnel mainly associated with the Saas- and Muttenglacier as weIl
as the Leckihorn area at elevations of 2800-2900 m a.s.l. Proceeding to the south
along the Bedretto-tunnel an additional recharge area along the Saashrner slope
at elevations of 2500-2600 m a.s.l. is drained towards the Bedretto-tunnel. Along
the fault zone seetion of the gallery, depleted
18
0 values diverging from the above
trend are observed. Arecharge altitude of rv2740 m a.s.l. is approximated from
the isotope data. The delineated recharge area for the fault zone section from
the model simulations with distributed recharge rates is situated at the glaciated
Witenwasserenpass area at an elevation of rv2800 m a.s.l. Additionally an area
along the northern slope of the Rotondo at an elevation of 2800-2900 m a.s.l. is
delineated as recharge area to the fault zone section. For the southern end 6"
18
0
values similar to the most northern end of the gallery are observed. The model
simulations delineate arecharge area along the upper slopes of the Gerenglacier
as recharge source to this tunnel section. A decreasing recharge altitude along
the northern end of the tunnel is also reproduced in the model variants with
uniform recharge rates. To the west of the Bedretto-tunnel arecharge area along
the Saashrner is delineated with altitudes between 2800 and 2700 m a.s.l. and
to the east of the gallery between 2900 and 2700 m a.s.l. as sampling stations
in the tunnel advance southwards. The fault zone section is recharged from the
Witenwasseren- and Gerenglacier at elevations ranging from 2700-2800 m a.s.l.
to the east and 2800-2900 m a.s.l. along the Rotondo slope. Even though the
range of values deduced from both recharge distributions show an agreement it
has to be stressed that the specific position and thus altitude of the recharge
areas is subject to the prescribed orientation of k
1
within the Rotondo granite,
which was constrained by field observations and measurements.
67
The chemical composition of the ground water in the Bedretto-tunnel and the
analysis of 34Sp80 content in aqueous S O ~ - of the sampIe waters indicate that
the sampled ground water is derived from within the granite body. Yet a poten-
tial ground water contribution from the northern orthogneiss and in particular
from local amphibolite wedges could serve as an explanation for the alignment
of the 34Sp80 data long a mixing-line of slope ,,-,0.6 and the variation in Ca/Na
ratio. One of these local amphibolite wedges is situated within the Leckihorn
area and the particle tracking for the model scenarios with distributed recharge
rates indicate a potential minor contribution of ground water originating in this
area to the inflow to the Bedretto-tunnel.
Estimated mean residence times from lumped-parameter modeling of the
18
0
content in ground water sampIes taken along the fault zone section are 1-1.5
years. In the numerical simulations, the advective travel times deduced from
the ca1culated isochrones far pathlines to the fault section of the gallery are in
a similar order of magnitude, however showing a wider range from 1-5 years
depending on the applied flow porosity value. Travel times to the fault zone
section of the Bedretto-tunnel were furthermore estimated on the basis of tritium
data in ground water sampIes and recharge components (Appendix D) using the
decay equation (Clark and Fritz 1997)
(3.7)
with a ~ H initial tritium concentration in ground water recharge, a ~ H residual
tritium concentration in ground water sampIe (both in TU) and A decay term
A = ln2/t
1
/
2
(t1/2 = 12.43 years, half-life of tritium). These simplified ca1cula-
tions underline that for the above range of travel times, summer precipitation is
only a minor component in ground water recharge to the fault zone section of the
Bedretto-tunnel.
The tritium data measured along the Bedretto-tunnel shows, that the ground
water encountered along the fault zone section is recharged from glacial meltwater
as a combination of meltwater from accumulated low tritium winter precipitation
and glacial icemelt. The contribution of submodern glacial icemelt thus leads to
68
the observed low tritium values in the inflows to the gallery. All model scenarios
and variants reproduce the observations of glacial recharge.
To the northern end of the Bedretto-tunnel, measured tritium values indicate
recharge from recent winter precipitation (Figure 3.24. The model scenarios with
distributed recharge rates delineate the Saas- and Muttenglacier as an impor-
tant recharge source to this tunnel section while the model variants with uniform
recharge delineate the slopes of the Saashrner as recharge area. If the Saas-
and Muttenglacier were to contribute dominantly to the ground water recharge
to this tunnel section, the difference between the measured tritium values in this
section to the observed low tritium values along the fault zone section would need
to be explained by a different isotopic composition of the meltwater from theses
glaciers, as both tunnel sections would then be recharged essentially by the same
type of recharge source. Ground water recharge from recent accumulated winter
precipitation on the slopes of the Saashrner could however explain well the mea-
sured tritium values in the northern tunnel section. The tritium data observed
along the northern section of the Bedretto-tunnel is thus not in agreement with
the recharge sources delineated from the simulations with the distributed recharge
rates. These predict the upper regions of the glaciers as dominant recharge areas
to the tunnel with high recharge rates.
The dominance of these recharge areas results from the bi-modal ground water
recharge concept for the glaciated areas applied in the hydrological model. In this
concept, meltwater from above the equilibrium line of the glaciers preferentially
contributes to the slow runoff storage and thus to ground water recharge, while
meltwater from below the equilibrium line is attributed to the fast runoff storage.
This model has been previously applied for hydrological studies in the Rhne-area
(Badoux 1999) and builds on the idea that meltwater from above the equilibrium
line percolates through the glacial body and crevasses towards the distributed
subglacial drainage system. Here its contact time with the fractured bedrock
surface is potentially larger than for meltwater originating c10ser to the glacier
snout. In these regions of the glacier a higher portion of the meltwater runs
off via the glacier surface and through a more developed channelized subglacial
drainage system (Tranter et al. 1996). However, this concept is based on strong
generalized assumptions, especially as the geometry of the subglacial drainage
system within the glacier is highly variable and no constrains, as to in which
69
areas below the glaciers of the study area ground water recharge may occur, are
available. Additionally, meltwater from the glacial ice is available for ground
water recharge throughout the base of the temperate glaciers due to frictional
heat and the conductive heat flux from the bedrock (Menzies 1995). Therefore,
the assumption of the equilibrium line as a limit for preferential ground water
recharge from the glaciers poses a strong constrain on the spatial distribution of
the recharge in these regions. Furthermore, the high amount of predicted ground
water recharge in these small regions of the glaciers is striking (up to 40% of mean
precipitation rate). Even though substantial recharge may occur along fractures
(Lerner et al. 1990) with considerable water intake rates (Rasmussen and Evans
1993), such localized high recharge rates in the upper regions ofthe glaciers seem
less plausible. A comparable recharge flux, distributed over a wider area of the
glaciated regions might be a better approximation of the ground water recharge
below the glaciated areas.
A second factor for the dominance of the localized high recharge areas in the
glaciated regions lies in the strong contrast in recharge rates between these areas
and the neighboring steep bare rock slopes, were very little recharge is predicted
by the hydrological model. According to Forster and Smith (1988a), recharge
rates along the summits might reach ::;10% of the precipitation rate. While it is
plausible that liquid precipitation will preferentially enter fast surface runoff along
the steep slopes, meltwater from the winter snowcover is an important potential
recharge source at these altitudes. This has been described by Rodhe (1998) as
a very efficient input to ground water recharge. Snowcover along the Saashrner
for example is generally substantial and especially topographie depressions, in
this case small kars at the foot of the steep rockfaces, might serve as potential
recharge areas (Lissey 1968; Lerner et al. 1990). These numerous small areas are
however not distinguished in the hydrological model due to the resolution of the
applied model grid. The above mentioned considerations might thus lead to an
overestimation of the contrast in recharge rates along the partly glaciated crests
of the mountain ridges in the Rotondo area.
The spread of the encountered tritium data in the northern section of the
gallery has been attributed to varying flow velocities (reflected by different tri-
tium concentrations) within separate hydraulic features such as fractures and
70
small scale faults. In this sense, the ground water system might be viewed as an
assemblage of discrete water parcels which move along defined flow paths (Gat
1981). Similar observations were made by Fontes et al. (1981) in a comprehensive
study in the Mont-Blanc tunnel. Even though the homogeneous continuum ap-
proach can not reproduce the potential heterogeneity of flow velocities present in
the granite body due to discrete small scale fractures and faults a certain spread
of deduced travel times for flow towards this northern tunnel section can also be
observed in this approach (Table 3.6).
The tritium data at the southern end of the accessible part of the Bedretto-
tunnel displays a similar discrepancy with regard to the recharge source as the
data at the northern end of the gallery. While the scenarios with a distributed
recharge predict mainly glacial meltwater as recharge source, the uniform recharge
variants delineate the slopes of the Wittenwasserenstock and the Rotondo as main
recharge sources with only a minor glacial component.
In general the findings from the analysis of the chemical and environmental
isotope composition of the ground water encountered along the Bedretto-tunnel
are reproduced in the model simulations. In particular the analysis of the ob-
served tritium data along the Bedretto-tunnel allows an evaluation of the plau-
sibility of the different recharge conditions applied in the model simulation.
3.7 Conclusion
In this paper we presented hydraulic measurements from a deep tunnel in frac-
tured crystalline rocks of the Swiss Alps and used this data subsequently to
constrain a regional ground water flow model. In this model we investigated the
effects of controlling factors on the flow system such as recharge conditions and
structural elements in model simulations. The results of these simulations were
compared to the findings of a previous study, where the environmental isotope
composition and the hydrochemistry of the ground water was analyzed. In the
framework of the chosen continuum approach, the numerical simulations were
able to provide valuable insights on the impact of the subsurface galleries on the
natural flow field within the Rotondo area and increased our understanding of
the most influential parameters affecting the position of the ground water table
71
in a high alpine catchment area. The water table position was shown to be sen-
sitive to even small changes in prescribed hydraulic conductivities, with changes
in elevation of several hundred meters for modifications of hydraulic conductivity
of less than one order of magnitude. The significant decline of the water table
below the mountain ridges of up to 600 m observed in the simulations is enhanced
by the presence of the subsurface galleries, which generally modify the natural
flow field by enhancing natural head gradients or even causing areversal of the
regional head gradients. The presence of a higher conductive fault zone with a
k-value contrast to the surrounding bedrock of 1:10 causes no significant impact
on the regional ground water table but most importantly acts as fast flowpath
for ground water recharging in glaciated areas in the eastern part of the study
area and discharging to the Bedretto-tunnel. The internal anisotropy ratio of
the horizontal components of hydraulic conductivity is less important due to the
overall sub-parallel orientation of the fault zone with the orientation of the major
horizontal component of hydraulic conductivity within the adjacent granite. Ca-
librated k-values for the Rotondo granite in the order of6.5x10-
9
to 1.2x10-
8
lie
in the range of reported equivalent k-values for Hercynian granites of the Central
Alps (Loew et al. 1996).
Spatially distributed ground water recharge rates significantly affect the head
distribution within the research area when compared to simulations with uni-
form recharge estimates. High relief massifs are more sensitive to changes in the
recharge rate with regard to the water table elevation.
Compared to the interpretations based on the analysis of environmental iso-
topes and hydrochemistry, the numerical simulations underline that the ground
water recharge to the Bedretto-tunnel is derived from within the granite body
with a potentially minor recharge contribution from an isolated amphibolite
wedge in the Leckihorn area towards the northern end of the gallery.
In the combined interpretation of environmental isotopes and hydrodynami-
cal modeling we found the tritium data a very useful tool in the delineation of
ground water recharge sources in this alpine region. The analysis has shown that
while the fault zone section in the Bedretto-tunnel is recharged by the glaciated
areas to the east of the gallery, the northern section of the tunnel receives mainly
ground water inflow recharged from recent precipitation with a major compo-
nent of accumulated winter precipitation. In the comparison of these findings
72
we could show that the spatially distributed recharge rates extracted from the
hydrological model overestiIrmte the contribution of the glaciated areas in ground
water recharge to the northern part of the tunnel. This is linked to the simplified
assumption made in the hydrological model concerning the spatial discretization
of ground water recharge in the glaciated areas. A more physical based model of
the glaciated areas (e.g. Arnold et al. (1998), Richards et al. (1996)), especially
of the meltwater routing through the glacial body would aid the better under-
standing of the spatial availability of potential ground water recharge and thus
lead to a better representation of theses recharge sources within the numerical
ground water flow model.
Acknowledgements
Access to the Furka-basetunnel and the Bedretto-tunnel and logistic help was
kindly provided by the Furka-Oberalp-Bahn. The digital elevation map DHM25
of the research area was used with approval by the Bundesamt fr Landesto-
pographie, Wabern (eH; Reference: BA013094). This research was funded by
the ETH research fund.
73
Table 3.1: Mean orientation offracture sets with precision k and apical half-angle
() of 95%-confidence cone from Fisher-analysis together with corresponding mean
normal fracture frequency A (and standard deviation a; Priest 1993) from surface
and tunnel outcrops.
Number of set
1
2
3
4
Mean orientation k
()
A
SUT
face ((J) Atunnel ((J)
strike/dip [.]
[0] [rn-I] [rn-I]
049/75 SE 8.8 7.0 2.1 (2.2) 0.5 (0.2)
080/83 SE 16.5 5.0 1.5 (1.2) 0.2 (0.3)
140/86 SW 20.7 3.5 0.7 (0.6) 2.6 (2.7)
170/79 SW 7.4 5.8 0.04 (0.07) 0.3 (0.2)
74
Table 3.2: Mean measured integral inflow rate to the accessible section of the
Bedretto-tunnel (total) and to the fault zone section within the Bedretto-tunnel
together with integral inflow rate to the Rotondo granite (RG) section of the
Furka-basetunnel as wen as to the whole Furka-basetunnel (measured at the Por-
tals). Further tabulated are standard deviation (J as indicator for variability
over time and associated error of the measurement (t only four measurements
November 98 - March 99). For tunnel metrics refer to Figure 3.4.
Tunnel section Tunnel metries Q (J Measurement error
[m] [lj8] [1/8] [lj8]
Bedretto-tunnel TM3780-5218 14.9 0.3 0.5
Fault zone TM4200-4320 7.5 0.1 0.5
Furka-ba8etunnel (RG) BM4900-8050 25.8 1.2 5.6
PortalOberwald BM-11400 94.4 1.7 5.9
Portal Realpt BM11400-15400 75.8 0.1 5.9
Table 3.3: Analytical approximations of hydraulic conductivities. Values are
ca1culated for the fault zone section and the remaining granite section as wen as
for the total tunnel section inc1uding the fault zone section
Section of the Bedretto-tunnel
Fault zone
Remaining tunnel section
Total tunnel section
Hydraulic conductivity (m/s)
6 . 0 ~ 8 . 2 X 10-
8
6.0-8.3 X 10-
9
1.0-1.4X 10-
8
75
Table 3.4: Initial approximations of effective hydraulic conductivities (hc) for the
geological units within the model domain together with calibrated hydraulic con-
ductivities for the individual scenarios (nc: no change; s': southern; n': northern)
Name Index Initial hc [m/s] Calibrated hc [m/s]
Scenario 1 Scenario 2 Scenario 3
Rotondo granite RG
1.1x 10-
8
1.2x10-
8
6.0x 10-
9
6.5x10-
9
Aar granite A 3.8x 10-
9
nc nc nc
s' Aar gneiss zone AK 5.3x 10-
9
nc nc nc
Urseren-Gavera- UGZ 1.5x 10-
8
nc nc nc
zone
n' paragneiss PN 3.1 x 10-
8
3.1 X 10-
9
3.1 X 10-
9
3.1 X 10-
9
Orthogneiss 0 2.3x 10-
8
9.2x10-
9
l.4x 10-
8
1.4xlO-
8
s' paragneiss P 3.1x 10-
8
nc nc nc
Ultra-helvetic M 4.0xlO-
7
nc nc nc
sediments
Fibbia granite F 1.1x 10-
8
nc nc nc
Cacciola granite C
1.1xlO-
8
8.0x10-
8
8.0xlO-
8
8.0xlO-
8
Fault zone FZ 1.2x 10-
7
6.0xlO-
8
6.5 X 10-
8
76
Table 3.5: Calibration targets for inflow rates along specific tunnel sections and
hydraulic heads in monitoring intervals with calibrated values for the individual
model scenarios Uonly four measurements November 1998 - March 1999)
Tunnel inflow rate
Tunnel Calibration Scenario 1 Scenario 2 Scenario 3
section target [ljs] [ljs] [ljs] [ljs]
Bedretto 14.9 0.5 14.8 15.0 15.2
Fault zone 7.5 0.5 (1.9) 7.7 7.8
Furka (RG) 25.8 5.6 25.2 21.8 21.8
Oberwald 94.4 5.9 95.0 91.6 89.7
Realpt 75.8 5.9 70.9 72.9 73.2
Hydraulic heads
Monitoring Calibration Scenario 1 Scenario 2 Scenario 3
interval target [m] [m] [m] [m]
Interval 1 1988 1819 1851 1850
Interval 2 1894 1756 1782 1781
77
Table 3.6: Approximated advective travel times across recharge areas for model
scenarios, deduced from ca1culated isochrones along pathlines initiated at the
Bedretto-tunnel nodes
Recharge area Travel time [years]
nfRG = 2.0 x 10-
4
nfRG = 1.0 x 10-
3
Leckihorn 2-3 8-10
Witenwassernstock 2-3.5 10-14
Rotondo 2-2.5 10-12
Saashrner 2-2.5 9-10
Witenwasserenpass 1-1.5 3-5
(through fault zone)
78
Table 3.7: Simulated ground water inflows to the tunnel galleries and simulated
heads in the monitoring intervals far model variants with uniform ground wa-
ter recharge rates (Variant 1 = 5.75x 10-
4
m/d, variant 2 = 12.05x 10-
4
m/d, vari-
ant 3 = 2.74x 10-
4
m/d).
Tunnel inflow rate
Tunnel section Variant 1 [1/s] Variant 2 [l/s] Variant 3 [l/s]
Bedretto 14.7 15.1 13.6
Fault zone 7.5 7.6 7.0
Furka (RG) 19.1 18.3 14.8
Oberwald 86.4 86.7 65.0
Realp 69.2 72.6 64.5
Hydraulic heads
Monitoring Variant 1 [m] Variant 2 [m] Variant 3 [m]
interval
Interval 1 1847.0 1854.6 1774.5
Interval 2 1779.1 1784.8 1719.1
79
4 kilometers
I
2 o
I
~ Geological boundary
,/ - ./ Riverlstream
J!k,. Summit (Ro: Piz Rotondo) Tunnel
Viliage (Re:Realp, Ow:Oberwald, Ro:Ronco) IK] Geological unit
Figure 3.1: Schematic plan view of the model domain indicating the projected
trace of the subsurface galleries and the individual geologieal units of the Aar-
and Gotthard-Massif (AM and GM) within the domain (A Aar-granite (AM);
AK southern gneiss zone (AM); UGZ Urseren-Gavera-zone; PN northern parag-
neiss (GM); 0 orthogneiss (GM); Pr southern paragneiss (GM); M ultra-helvetie
sediments (GM); RG Rotondo granite (GM); C Caeeiola granite (GM), F Fibbia
granite (GM)).
80
.2904 Summit with altitude
me and identification
me Mettligrat, ro Rotondo, wi Witenwasserenstock,
le Leckihorn, rg Rotlligrat, sa Saashrner,
mu Muttenhrner, st Stellibodenhorn, pe Pesciora
Glaciated area
gg Gerenglacier
wg Witenwasserenglacier
sg Saasglacier
mg Muttenglacier
Spring discharge
Elevation contour interval =75 m
Figure 3.2: Enlarged plan view of the upper Gerental-valley in the model domain.
Indicated are the projected tunnel trace together with the position of the glaciers
and the trace of the model fault zone intersecting the Bedretto-tunnel.
52
.Change of sensor set-p
l
50
48
';::'
(ll
B
l!!
::l
46
<n
<n
l!!
a.
...!.
(ll
44
~
42
40
38
81
1.10.1999 1.01.2000 1.04.2000
Date
1.07.2000 1.10.2000
Figure 3.3: Long-term record of hydraulic pressures In the research borehole;
data from monitoring interval 1 and 2.
OJ
~
o
OJ
~
.,.
CD
o
o
TMO
(,0
- o ~
direction of tunnel drainage
Figure 3.4: Schematic trace of the Furka-basetunnel and Bedretto-tunnel with
indicators for the direction of tunnel drainage and tunnel metrics. Metrics refer
to meter from the Oberwald- and Ronco-portal respectively.
82
0-2
3-7
8 -12
13 - 15
16 - 19
_ 20-27
_ 28-32
_ 33-37
_ 38-42
x 10-4 m/d
o
I
5km
I
Figure 3.5: Map view of the distributed ground water recharge rates applied in
the ground water flow model.
83
D 0-2
D 3-7
19i1lJ 8 - 12 AN
_ 13-15
_ 16-19
_ 20-27
.. 28- 32 Tunnel
_ 33 - 37 ... Summ;'
_ 38-42
x10-
4
m/d
500m
,
Figure 3.6: Enlarged map view of the distributed ground water recharge rates in
the upper Gerental-valley, applied in the ground water flow model.
Topography
Topography -100 m
Topography -200 m
Topography -700 m
Tunnel elevation +100 m
Tunnel elevation
Tunnel elevation -5 m
450 m a.s.I.
oma.s.1.
84
Layer 1
Layer 2
Layer3
Layer 4
Layer 5
Layer 6
Layer 7
Layer 8
Layer 9
Layer 10
Layer 11
Layer 12
Layer 13
Slice 1
Slice 2
Slice 3
Slice 4
Slice 5
Slice 6
Slice 7
Slice 8
Slice 9
Slice 10
Slice 11
Slice 12
Slice 13
Slice 14
Figure 3.7: Schematic overview of the layer structure applied In the 3d finite
element model
85
o 6km
L..' --',
Figure 3.8: 3d view of finite element mesh
86
NW
5100
scenario 2
, , scenario 3
""'-. measured
.........
..... scenario 1
4300 4700
Tunnel metries [m]
3900
:-... -- .... :-: .................
.......... '. --.::-:
. ~
.....~ .
"' ' ~
......
. ~
'" .
", ......
., .....
\)
\ ~
. ~
.
.
.,
......... .......:
SE
1000
800
c
E
=..
:::
600
0
Cj::
c
Q)
>
~
400
:::l
E
:::l
0
200
0
3500
Figure 3.9: Plot of measured cumulative ground water inflow to the Bedretto-
tunnel along the tunnel profile versus simulation results for the individual model
scenarios. The whole section is situated within the Rotondo granite. For tunnel
metrics refer to Figure 3.4.
87
Topography, contour interval = 80 m
~ Depth to water table below ground surface, contour interval 100 m
o
I
5 km
I
.... Mountain summit
__ Tunnel galleries
Figure 3.10: Depth to the ground water table below ground surface for model
scenario 1. Also indicated are several mountain summits far better orientation.
88
Topography, contour interval =80 m
~ Depth to water table below ground surface, contour interval 50 m
Mountain summit
___ Tunnel galleries
scenario 2. Also indicated are several mountain summits for better orientation.
Glaciated area
mg Muttenglacier
sg Saasglacier
wg Witenwasserenglacier
gg Gerenglacier
89
./ Tunnel gallery
4'" Recharge area for/model scenadmt
, ~ (distributed recharge rate)
......A'f17 Recharge area for kooel vsriaffi-
4J.jJY !t & Z(uniform recharge rate)
.... Summit
mu Muttenhorn
le Leckihorn
sa Saashrner
wi Witenwasser-
enstock
ro Rotondo
Figure 3.12: Enlarged plan view of the upper Gerental-valley with the position
of the recharge areas to the Bedretto-tunnel for the model scenarios and model
variant 1&2.
90
1400r-----------------------,
1200
........
.............
!i I
i 400 . .
:it
200 .
3700 3900 4100 4700 4500 4300
Tunnel Melries [m]
4900 5100

5300
Figure 3.13: Range of advective travel times to the Bedretto-tunnel for 120m in-
tervals along the tunnel axis. Travel times are deduced from calculated isochrones
along pathlines initiated at the Bedretto-tunnel nodes (njRG = 2.0 x 10-
4
)
6000r- ---,
5000
w
iU' 4000
"E
.,
" 3000
"
.,
U 2000
"
-i5
..:
1000
...... 11
ILIIIUr .
...........................................................................I .
0-1-- --_--_-_--_--_----4
5300 5100 4900 4700 4500 4300 4100 3900 3700
Tunnel Melries [m]
Figure 3.14: Range of advective travel times to the Bedretto-tunnel far 120m in-
tervals along the tunnel axis. Travel times are deduced from calculated isochrones
along pathlines initiated at the Bedretto-tunnel nodes (njRG = 1.0 x 10-
3
)
91
~ Drawdown due to tunnel galleries, contour interval 50 m
o
I
5 km
I
... Mountain summit
__ Tunnel galleries
Figure 3.15: Plan view of the model domain with isolines of drawdown due to
the tunnel galleries for model scenario 1. Also indicated are several mountain
summits for better orientation.
WSW
92
ENE
.---------------------,------r12
x 10-
4
m1d
8
4
(a)
'-- .::.- -J.- -"'-O
b
(b)
(c)
1- - water table I
2560
2930fn
F-B Furka-basetunel
B-T Bedretto-tunnel
sa Saashoerner
Figure 3.16: 2d section of hydraulic potentials for model scenario 3. a) Ground-
water recharge rates extracted along the profile section and interpolated on a 50
m interval along the section; b) 2d-section of hydraulic potentials and position of
ground water table for model scenario 3 simulated with the tunnels and c) corre-
sponding section far model scenario 3 simulated without the tunnels; d) schematic
plan view of the location of the profile section in the upper Gerental-valley (refer
to Figure 3.2). The vertical exaggeration of b) and c) is 0.7.
93
o
I
5 km
I
Mountain summit
~ Depth to water table below ground surface, contour interval 100 m __ Tunnel galleries
variant 1 with uniform ground water recharge of 5.75 X 10-
4
mfd. Also indicated
are several mountain summits for better orientation.
94
o
I
5 km
I
... Mountain summit
~ Depth to water table below ground surface, contour interval 100 m Tunnel galleries
4
m/d. Also indicated
are several mountain summits for better orientation.
i
o
E
(b)
,
o
WNW
(c)
95
,
6630m
W
,
5825m
ESE
Figure 3.19: a) and b) 2d section of hydraulic potentials for model scenario 3
(without tunnel condition) indicating regional and local ground water flow sys-
tems. c) Location of sections within the model domain. The vertical exaggeration
of a) is 1.4 and of b) is 1.2
96
I
1900m
ESE
--, /-
'" .... "1
'1
_--I
f
I
\
I
/1
" , , ~
- \. '- /
" " ' ~
, \/,\ '"
(a)/(b) fz
I
o
WNW
(c)
Figure 3.20: 2d section of hydraulic potentials for model scenario 2 (a) and
scenario 1 (b) (without tunnel condition) indicating the influence of the large
scale fault zone on the head distribution in the Saashorn area. c) Location of
section and fault zone (fz) within the model domain. Vertical exaggeration of a)
and b) is 0.5
97
I
o
WNW
(cl
2550
! ' ~ ~ _ .... -
--''-, J
,/ .... -/
r
"
__ J:
I
I
I
2815m
ESE
Figure 3.21: 2d section of hydraulic potentials far model scenario 2 (a) and
scenario 1 (b) (without tunnel condition) indicating the influence of the large
scale fault zone on the head distribution in the Leckihorn area. c) Location of
section and fault zone (fz) within the model domain. Vertical exaggeration of a)
and b) is 0.5
98
0.50
045
040
0.35
0 0.30
Cf)
I-
N
0.25
11
""0
er 0.20
hd = 10
hd = 100
0.15
0.10
0.05
hd = 1000
hd = 10000
._._.- _..
_._._.- --1
hd = 100000

104
Figure 3.22: Dimensionless flow rate qd versus dimensionless time td for constant
head test with a finite radius wen and a constant head boundary at varying dimen-
sionless distances h
d
according to the analytical solution of Murdoch and Franeo
(1994). Also depicted are the asymptotic values ql (dashed lines) representing
steady state conditions.
20
18
16
14
12
;R
0
10
""C
er
<]
8
6
4
2
99
10
9
1010 10
11
td =(Tt)/(r
2
w
S)
Figure 3.23: Residual percentage between dimensionless flowrate qd and asymp-
totic value of dimensionless flowrate ql representing steady-state condition versus
dimensionless time td far constant head test in a finite radius wen with a constant
head boundary at a dimensionless distance of h
d
= 800
SE
100
NW
e......
5'
t::..
E
:::J
:;:::;
o
o
co
I<:>
3200
0 500 m
3000
.. sampled inflow
2800

2600
..-.
2400
Tunnel collapsed
1480
__________ __ __"'_""__ !"9__ ! __
22.5
[]
20.0

..

17.5
15.0

12.5
16
llil
t
12 f tt
t
f
f
8
f
4
fl
f
0
-14.6
@]
11
-14.8
11
1111
1
-15.0
11
-15.2
Figure 3.24: Cross-section along the profile of the Bedretto-tunnel with location
of sampled inflows; a) mean temperature, b) mean tritium concentration and c)
mean 0
18
0 value of sampled ground water. The analytical error is indicated.
4
Concluding remarks
The hydrochemical and isotopic investigations presented in paper 1 allowed us to
assess the ground water encountered along the Bedretto-tunnel to be essentially
of recent meteoric origin. Discharge rates, chemical and isotopic composition
show only subtle temporal variations and even though this finding is not surpris-
ing considering the depth of the subsurface gallery, the measurements also c1early
indicate the absence of higher mineralized older ground water components. Fur-
thermore the data shows that amongst the subtle variations, the highest fluctu-
ations can be found in the ground water inflow along the fault zone section.
The chemical composition of the ground water indicates an evolution of the
composition along the flow path within the granite body governed mainly by
the process of feldspar hydrolysis. The c5
18
0-data supports this interpretation,
underlining the meteoric origin of the ground water. 34Sp80 data is generally in
agreement with this interpretation. Yet, in the northern part of the tunnel, the
data 34Sp80 data indicates a potential contribution of ground water originating
in the neighboring gneiss formation.
Tritium and c5
18
0 data of ground water sampies reflect the significant con-
tribution of accumulated recent winter precipitation to ground water recharge.
In particular, low tritium values measured in sampies along the fault zone sec-
tion of the tunnel show the imprint of glacial meltwater to ground water recharge.
Lumped-parameter modeling of the c5
18
0 signal lead here to mean residence times
of 1-1.5 years implying rather rapid recharge via the fault zone to the gallery.
Even though simplifications needed to be applied in the analysis, the collected
database provided valuable indicators for ground wate recharge conditions and
101
102
the significance of the fault zone as preferential flow path.
The numerical simulations presented in paper 2 allowed us to study the signif-
icance of individual parameters on the behaviour of the flow system and provided
some additional confidence in the interpretations proposed in the first paper. The
simulations illustrate the impact of the subsurface galleries on the ground wa-
ter flow system, causing significant additional drawdown of the water table and
leading to strong modifications of the head gradients in the study area. The
presence of the higher conductive fault zone causes no significant impact on the
regional ground water table but most importantly is shown to act as fast flowpath
for ground water recharge from the glaciated areas towards the Bedretto-tunnel.
This is illustrated by the estimated advective travel times of 1-5 years for flow
from the glaciated areas via the fault zone to the Bedretto-tunnel.
The spatial distributed recharge rates, with a high contrast in recharge rates
along the mountain ridges between glaciated areas and bare rock slopes, strongly
affect the water table elevation compared to uniform recharge rate model variants.
The simulations with varying uniform recharge rates illustrated the non-linear
response of the water table elevation on changes in ground water recharge rates.
The model simulations increased our confidence in the interpretation of glacial
meltwater contribution to ground water recharge via the fault zone to the Bedretto-
tunnel, as consistently all model scenarios and variants delineate the glaciated
areas east of the tunnel trace as recharge areas to the fault zone section of the
gallery.
The inability of the flux-calibrated model to reproduce the measured local
hydraulic head data in the research borehole indicates that a match of both
fluxes and local head measurements can not be achieved with the presented ap-
proach investigating the effect of a single large scale model fault zone. For this,
additional higher conductive features would need to be introduced into the model.
Comparing the findings from the hydrochemistry/isotope study (paper 1) to
the results of the numerical simulations we see that the simulations with dis-
tributed recharge rates delineate recharge areas in the north and west of the
Bedretto-tunnel which do not agree with the measured tritium signal in the north-
ern part of the Bedretto-tunnel. Here the tritium measurements do not indicate a
103
contribution of low tritium glacial meltwater to ground water recharge as it is ob-
served along the fault zone section. Yet the simulations delineate essentially the
Saas- and Muttenglacier as main recharge areas. Simulations with comparable
uniform recharge rates however are in agreement with the observed tritium signal,
as recharge areas are then aligned along the mountain slopes where accumulated
recent winter precipitation as major recharge source would lead to the observed
tritium signal. This discrepancy indicates the impact of the strong simplifications
made in the hydrological model for the discretisation of the recharge rates in the
glaciated areas. A more detailed physically based model of the glaciated areas as
mentioned in paper 2 would probably be able to avoid this ambiguity and lead
to a better representation of these recharge sources within the numerical ground
water flow model.
In all model scenarios and variants the glacial meltwater component in ground
water recharge to the Bedretto-tunnel via the fault zone is reproduced. Mean
residence times deduced from the 0
18
0 data of 1-1.5 years are in agreement with
the order of magnitude in the estimated advective travel times from the numerical
simulations of 1-5 years for flow from the glaciated areas via the fault zone to
the Bedretto-tunnel. The observed recharge altitude dependency in 0
18
0 values
along the Bedretto-tunnel is also reproduced by the numerical simulations.
Minor ground water contributions from the northern gneiss-formation (0) de-
scribed in paper 1 as possible explanation for the alignment of the 34Sp80 data
along a mixing line are reproduced by the numerical simulations with distributed
recharge rates.
In retrospect, the combination of the results from the hydrological model, i.e.
the distributed ground water recharge rates, with additional analysis of the hydro-
chemistry and environmental isotope content of the ground water and hydraulic
measurements has shown to be a valuable approach to increase our understand-
ing of the behaviour and the influencing factors of the groundwater flow systems
in the fractured Rotondo granite. In particular the analysis of the environmen-
tal isotopes provided a useful tool as indicators for recharge conditions and as
a basis to evaluate the results of the numerical simulations. In this context the
tritium data has shown to be valuable in assessing the plausibility of the applied
distributed recharge rates.
104
The numerical simulations show some short-falls in terms of a comprehensive
representation of the natural conditions as indicated by the inability to correctly
reproduce the measured local head data. However, in the framework of the
presented approach we were able to demonstrate the influence of individual con-
trolling factors on the flow system in different model scenarios and also gained
additional confidence in the interpretations made on the basis of the measured
isotope data.
References
Abbott, M., A. Lini, and P. Bierman (2000). 6
18
0 and 3H measurements
constrain groundwater recharge patterns in an upland fractured bedrock
aquifer, vermont, USA. Journal of Hydrology 228, 101-112.
Abelin, H., L. Birgersson, L. Moreno, H. Widen, T. Agren, and I. Neretnieks
(1991). A large-scale flow and tracer experiment in granite - 2. Results and
interpretation. Water Resources Research 27(12), 3119-3135.
Ambach, W., H. Eisner, R. Haefeli, and M. Zobl (1971). Bestimmung von
Firnrcklagen am Eisschild Jungfraujoch durch Messung der Gesamtbeta-
aktivitt von Firnproben. Z. Gletscherkunde Glazialgeologie 7, 57-62.
Ambach, W., H. Eisner, H. Moser, W. Rauert, and W. Stichler (1971).
Ergebnisse von Isotopenmessungen am Gletscherbach des KesseIwanfern-
ers (tztaler Alpen). Ann. Meteorol. 5, 209-218.
Amberg Messtechnik AG (1994). Furkatunnel. Korrosionsdmmung/Belf-
tung. Technical Report R23/1O, Furka-Oberalp-Bahn AG, Brig (CH).
Andersson, J.-E., L. Ekman, R. Nordqvist, and A. Winberg (1991). Hydraulic
testing and modelling of a low-angle fracture zone at Finnsjn, Sweden.
Journal of Hydrology 126, 45-77.
Arnold, A. (1970). On the history of the Gotthard-massif (Central Alps,
Switzerland). Eclogae Geologicae Helvetiae 63, 29-30.
Arnold, N., K. Richards, I. Willis, and M. Sharp (1998). Initial results from a
distributed, physically based model of glacier hydrology. Hydrological Pro-
cesses 12(2),191-219.
Badoux, A. (1999). Untersuchung zur flchendifferenzierten Modellierung von
Abfluss und Schmelze in teilvergletscherten Einzugsgebieten. Master's the-
105
106
sis, Institute of Geography, ETH-Zurich.
Balderer, W., J. Fontes, J. Michelot, and D. Elmore (1987). Isotopic investi-
gations of the water-rock system in the deep crystalline rock of northern
Switzerland. In P. Fritz and S. Frape (Eds.), Saline Water and Gasses in
Crystalline Rocks, pp. 175-195. St. John's, Newfoundland: Geological As-
sociation of Canada. GAC special paper 33.
Balderer, W., F. Pearson, J. Fontes, J. Michelot, and S. Soreau (1990). 3
4
S
and
18
0 in S04 as indicators of the origin of deep groundwaters in north-
ern Switzerland. In A. Parriaux (Ed.), Memoires 22nd Congress of fAH,
Lausanne, pp. 364-373. IAH.
Balek, J. (1988). Groundwater recharge concepts. In I. Simmers (Ed.), Esti-
mation of Natural Groundwater Recharge, Dordrecht, pp. 3-11. NATO ASI
Series, Series C, Vol. 222: D. Reidel Publishing Company.
Ball, J. W. and D. K. Nordstrom (1992). User's manual for WATEQ4F, with
revised thermodynamic data base and test cases for ca1culating speciation
of major, trace, and redox elements in natural waters. Technical Report
Open-File Report 91-183, U.S. Geological Survey, Menlo Park, CA.
Barker, J. A. (1988). A generalized radial flow model for hydraulic tests in
fractured rock. Water Resources Research 24 (10), 1796-1804.
Bear, J. (1993). Modeling flow and contaminent transport in fractured rocks.
In J. Bear, C.-F. Tsang, and G. de Marsily (Eds.), Flow and Contaminent
Transport in Fractured Rock, Chapter 1, pp. 1-38. San Diego: Academic
Press.
Beavan, J., K. Evans, S. Mousa, and D. Simpson (1991). Estimating aquifer
parameters from analysis of forced fluctuations in weIl level: An example
from the Nubian formation near Aswan, Egypt - 2. Poroelastic properties.
Journal of Geophysical Research 96(B7), 12139-12160.
Bourdet, D. and A. C. Gringarten (1980). Determination of fissure volume and
block size in fractured reservoirs by type-curve analysis. In Proceedings 55th
SPE Annual Fall Technical Conference, Dallas (TX) , Richardson (TX).
Soc. Petr. Eng. paper no. 9293.
107
Brace, W. (1980). Permeability of crystalline and argillaceous rocks. Interna-
tional Journal 0/ Rock Mechanics '3 Mining Science. 17, 241-251.
Brace, W. (1984). Permeability of crystalline rocks: New in situ measurements.
Journal 0/ Geophysical Research 89(B6), 4327-4330.
Cacas, M., E. Ledoux, G. De Marsily, B. Tillie, A. Barbreau, E. Durand,
B. Feuga, and P. Peaudecerf (1990). Modeling fracture flow with a stochas-
tic discrete fracture network: Calibration and validation - 1. The flow
model. Water Resources Research 26(3),479-489.
Cacas, M., E. Ledoux, G. De Marsily, A. Barbreau, P. Calmels, B. Gaillard,
and R. Margritta (1990). Modeling fracture flow with a stochastic discrete
fracture network: Calibration and validation - 2. The transport model.
Water Resources Research 26(3),491-500.
Cacas, M. C., E. Ledoux, G. de Marsily, B. Tillie, A. Barbreau, P. Calmels,
B. Gaillard, R. Margrita, E. Durand, B. Feuga, and P. Peaudecerf (1990).
Flow and transport in fractured rocks: An in situ experiment in the fanay-
augeres mine and its interpretation with a discrete fracture network model.
In Memoires 0/ the 22nd Gongress 0/ IAH, Vol. XXII, Lausanne, pp. 13-37.
IAH.
Caine, J. S., J. P. Evans, and C. B. Forster (1996). Fault zone architecture and
permeability structure. Geology 24 (11),1025-1028.
Caine, J. S. and C. B. Forster (1999). Fault zone architecture and fluid flow:
Insights from field data and numerical modeling. In W. C. Haneberg, P. S.
Mozley, J. C. Moore, and L. B. Goodwin (Eds.), Faults and Subsur/ace
Fluid Flow in the Shallow Grust, pp. 101-127. Washington: American Geo-
physical Union. Geophysical Monograph 113.
Carrera, J. and J. Heredia (1987). Inverse problem of chalk river block, HY-
DROCOIN level 2 (case3) and level 2 (case5a). Technical Report NTB-88-
14, NAGRA.
Carrera, J., J. Heredia, S. Vomvoris, and P. Hufschmied (1990). Fracture flow
modelling: Applieation of automatie calibration techniques to a small frac-
tured monzonitie gneiss block. In S. P. Neuman and I. Neretnieks (Eds.),
Hydrogeology 0/ Low Permeability Environments, pp. 115-167. IAH. Se-
lected papers on hydrogeology, Volume 2.
108
Cineo-Ley, H. and H.-Z. Meng (1988). Pressure transient analysis of wells with
finite eonduetivity vertieal fraetures in double porosity reservoirs. In Pro-
eeedings 63rd SPE Annual Teehnieal Conferenee, Houston (TX) , Riehard-
son (TX), pp. 645. Soe. Petr. Eng. paper no. 18172.
Clark,1. D. and P. Fritz (1997). Environmental Isotopes in Hydrogeology. Boea
Raton: Lewis Publishers.
Clauser, C. (1992). Permeability of erystalline roeks. EOS 73 (21), 233-240.
Coleneo (1993). Gotthard-Basistunnel: Quantitative Analyse hydrogeologis-
eher Aufnahmen von Stollen- und Tunnelbauten im Gebiet Aar-Massiv-
Gotthard-Massiv-Leventina. Teehnieal Report 1763/5, Coleneo Power Con-
sulting.
Craig, H. (1961). Isotopic variations in meteorie waters. Seienee 133, 1702-
1703.
Cruehet, M. (1985). Influenee de la deeompression sur le eomportement hy-
drogeologique des massifs eristallins en basse Maurienne (Savoie, Franee).
Geologie Alpine 61, 65-73.
Davison, C. and E. Kozak (1988). Hydrogeologie eharaeteristies of major frae-
ture zones in a large granite batholith of the Canadian shield. In Proeeedings
of the 4th Canadian/American Conferenee on Hydrogeology. Dublin, Ohio:
National Ground Water Assoeiation.
Dierseh, H.-J, G. (1998). FEFLOW - Interactive, Graphies-Based Fiite-
Element Simulation System for Modeling Groundwater Flow, Contaminant
Mass and Heat Transport Processes, Users's Manual. WASY.
Dole, M., G. Lange, D. Rudd, and D. Zaukelies (1954). Isotopie eomposition
of atmospherie oxygen and nitrogen. Geoehimiea et Cosmochimiea Acta 6,
65-78,
Dubois, J.-D. (1993). Typologie Des Aquijeres Du Cristallin: Exemple Des
Massifs Des Aiguilles Rouges et Du Mont-Blane (Franee, Italie et Suisse).
Ph. D. thesis, Eeole polyteehnique federale de Lausanne, Dep. Genie Civil.
Dverstorp, B. and J. Andersson (1989). Applieation of the diserete fracture
network eoneept with field data: Possibilities of model ealibration and val-
idation. Water ResouT'ees Research 25 (3), 540-550.
109
Evans, J. P., C. B. Forster, and J. V. Goddard (1997). Permeability of fault-
related rocks, and implications for hydraulic structure of fault zones. Jour-
nal 0/ Structural Geology 19(11), 1393-1404.
Evans, K., J. Beavan, D. Simpson, and S. Mousa (1991). Estimating aquifer
parameters from analysis of forced fluctuations in weH level: An example
from the Nubian formation near Aswan, Eqypt - 3. Diffusivity estimates for
saturated and unsaturated zones. Journal 0/ Geophysical Research 96(B7),
12161-12191.
Fisher, R. (1953). Dispersion on a sphere. Proceedings 0/ the Royal Society 0/
London A217, 295-305.
Flerchinger, G. N., K. R. Cooley, and D. R. Ralston (1992). Groundwater
response to snowmelt in a mountainous watershed. Journal 0/ Hydology 133,
293-311.
Fontes, J., G. Bortolami, and G. Zuppi (1981). Hydrologie isotopique
du massif du Mont-Blanc. In Isotope Hydrology, Vienna, pp. 411-440.
IAEA/UNESCO. Proc. Symposium Neuherberg, vol.l.
Fontes, J., D. Louvat, and J. Michelot (1986). Some constraints on geochem-
istry and environmental isotopes for the study of low fracture flows in crys-
talline rocks - the stripa case. In IAEA (Ed.), Isotope Techniques in the
Study 0/ the Hydrology 0/ Fractured and Fissured Rocks, Vienna, pp. 29-67.
IAEA.
Forster, C. and L. Smith (1988b). Groundwater flow systems in mountain-
ous terrain - 1. Numerical modeling technique. Water Resources Re-
search 24 (7),999-1010.
Forster, C. and L. Smith (1988a). Groundwater flow systems in mountainous
terrain - 2. Controlling factors. Water Resources Research 24 (7), 1011-
1023.
Forster, C. B. and J. P. Evans (1991). Hydrology ofthrust faults and crystalline
thrust sheets: Results of combined field and modeling studies. Geophysical
Research Letters 18(5),979-982.
Forster, C. B., J. V. Goddard, and J. P. Evans (1994). Permeability structure
of a thrust fault. Technical Report Open-File-Report 94-228, US geological
110
survey. Proceedings of workshop LXIII - The mechanical involvement of
fluids in faulting.
Frick, U. (1994). The Grimsel radionuclide migration experiment - a contribu-
tion to raising confidence in the validity of solute transport models used in
performance assessment. In GEOVAL '94 - Validation Through Model Test-
ing, Proceedings 01 an NEAjSKI Symposium, pp. 2 4 5 ~ 2 7 2 . OECD Nuclear
Energy Agency.
Frick, U., W. Alexander, B. Baeyens, P. Bossart, M. Bradbury, C. Bhler,
J. Eikenberg, T. Fierz, W. Heer, E. Hoehn, I. McKinley, and P. Smith
(1992). The radionuclide migration experiment - overview of investigations
1985-1990. Technical Report NTB-91-04, NAGRA.
Gat, J. (1981). Groundwater. In J. Gat and R. Gonfiantini (Eds.), Stable Iso-
tope Hydrology. Deuterium and Oxygen-18 in the Water Cycle, Number
210, pp. 223-240. Vienna: IAEA. Technical Report Series No. 210.
Gat, J. R. (2000). Atmospheric water balance - the isotopic perspective. Hy-
drological Processes 14, 1357-1369.
Goodman, R. (1965). Ground water inflows during tunnel driving. Engineering
Geology 2(2),39-56.
Gringarten, A. C., H. Ramey Jr., and R. Raghavan (1974). Unsteady-state
pressure distribution created by a weIl with a single infinite-conductive
vertical fracture. SPE Journal, 347-351. paper no. 4051.
Guimera, J. and J. Carrera (1997). On the interdependence of transport and
hydraulic parameters in low permeability fractured media. In T. Pointet
(Ed.), Hard Rock Hydrosystems, pp. 123-133. Wallingford: IAHS. IAHS
Publ. no. 241.
Gurtz, J., A. Baltensweiler, and H. Lang (1999). Spatially distributed
hydrotope-based modelling of evapotranspiration and runoff in mountain-
ous basins. Hydrological processes 13(17),2751-2768.
Gurtz, J., G. Peschke, and O. Mendel (1990). Hydrologie processes in small
experimental areas influenced by vegetation cover. In Hydrological Research
Basins and Environment, pp. 61-69. Proceedings and Information No.44,
Intern. Conf. Wageningen.
111
Hafner, S. (1958). Petrographie des sdwestlichen Gotthardmassivs. Schweiz-
erische Mineralogische und Petrographische Mitteilungen. 38, 255-362.
Hamm, S.-Y. and P. Bidaux (1996). Dual-porosity fractal models for transient
flow analysis in fissured rocks. Water Resources Research 32(9), 2733-2745.
Hatzseh, P. (1994). Bohrlochmessungen. Stuttgart: Ferdinand Enke Verlag.
Herbert, A., J. Gale, G. Lanyon, and M. R. (1991). Modelling for the Stripa
site characterization and validation drift - inflow: Prediction offlow through
fractured rock. Technical Report SKB91-35, SKB.
Himmelsbach, T., H. Htzl, and P. Maloszewski (1998). Solute transport pro-
cesses in a highly permeable fault zone of Lindau fractured rock test site.
Ground Water 36(5),792-800.
Hoefs, J. (1997). Stable Isotope Geochemistry. Berlin: Springer.
Horibe, Y., K. Shigehara, and T. Y. (1973). Isotope seperation factors of car-
bon dioxide-water system and isotopic composition of atmospheric oxygen.
Journal of Geophysical Research 78, 2625-2629.
Hsieh, P. A., S. P. Neuman, G. K. Stiles, and E. S. Simpson (1985). Field
determination of the three-dimensional hydraulic conductivity tensor of
anisotropie media - 2. Methodology and application to fractured rocks.
Water Resources Research 21 (11), 1667-1676.
ISO (1980). Water flow measurements in open channel using weirs and venturi
flumes - part 1: Thin-plate weirs. Technical Report ISO 1438/1-1980(E),
International Organization for Standardization, Geneve.
ISO (1997). Measurement ofliquid flow in open channels - velocity-area meth-
ods. Technical Report ISO 748:1997(E), International Organization for
Standardization, Geneve.
Jger, E. and E. Niggli (1964). Rubidium-Strontium-Isotopenanalysen an
Mineralien und Gesteinen des Rotondogranits und ihre geologische Inter-
pretation. Schweizerische Mineralogische und Petrographische Mitteilun-
gen. 44(1), 61-8I.
Jamier, D. (1975). Etude de la Fissuration, de L'hydrogeologie et de la Geo-
chemie Des Eaux Profondes Des Massifs de L'arpille et Du Mont-Blanc.
Ph. D. thesis, Faculte des Sciences de l'Universite de Neuchtel.
112
Jouzel, J. and L. Merlivat (1984). Deuterium and oxygen 18 in precipitation:
Modeling of the isotopic effects during snow formation. Journal of Geo-
physical Research 89 (NO D7), 11749-11757.
Kattelmann, Rand K. EIder (1991). Hydrologie characteristics and water
balance of an alpine basin in the sierra nevada. Water Resources Re-
search 27(7), 1553-1562.
Keller, F. and T. R Schneider (1982). Der Furka-Basistunnel - Zur Erffnung
am 25. Juni 1982; Geologie und Geotechnik. Schweizer Ingenieur und Ar-
chitekt 24, 512-520.
Keppler, A. (1996). Hydrogeologische, Hydrochemische und Isotopenhydrologis-
che Untersuchungen an den Oberflchen- und Kluftwssern im Grimselge-
biet, Schweiz. GSF-Bericht 4/96. Neuherberg: GSF-Forschungszentrum fr
Umwelt und Gesundheit.
Kimmermeier, E., P. Perrochet, A. R, and L. Kiraly(1985). Simulation par
modele mathematique des ecoulements souterrains entre les Alpes et la
Foret Noire. Technical Report NTB-84-50, NAGRA.
Kitterod, N.-O., H. Colleuille, W. Wong, and T. Pedersen (2000). Simulation
of groundwater drainage into a tunnel in fractured rock and numerical anal-
ysis of leakage remediation, Romeriksporten tunnel, Norway. Hydrogeology
Journal 8(5),480-493.
Klla, E. (1993). Gotthard-Basistunnel- Regionale Hydrogeologie im Projekt-
gebiet unter besonderer Bercksichtigung des Ritom-Gebietes. unpublished
report from 30.11.193.
Krouse, H. Rand B. Mayer (2000). Sulphur and oxygen isotopes in sulphate. In
P. G. Cook and A. L. Herczeg (Eds.), Environmental Tracers in Subsurface
Hydrology, pp. 195-231. Boston: Kluwer Academic Publishers.
Krouse, H. Rand R O. Van Everdingen (1986). Interpretation of oxygen iso-
tope data for sulphate in subsurface waters. In 5th International Symposium
on Water-Rock Interactions - Extended Abstracts, Reykjavik, pp. 663-666.
International Association of Geochemistry and Cosmochemistry.
Labhart, T. (1999). Aarmassiv, Gotthardmassiv und Tavetscher zwischenrnas-
siv: Aufbau und Entstehungsgeschichte. In S. Lw and R Wyss (Eds.), Vor-
113
erkundung und Prognose der Basistunnels am Gotthard und am Ltschberg,
Rotterdam, pp. 31-43. A.A. Balkema. Proceedings Symposium Geologie
Alptransit.
Legchenko, A., A. Beauce, U. Ofterdinger, and P. Renard (1998). Field tests
of the surface proton magnetic resonance instrument NUMIS in fractured
crystalline rocks of the western Gotthard-massif (Switzerland). Technical
Report BRGM-ETH3465/13, BRGM/ETH.
Lerner, D. N., A. S. Issar, and I. Simmers (1990). Groundwater Recharge -
A Giude to Understanding and Estimating Natural Recharge, Volume 8 of
International Contributions to Hydrogeology. Hannover: Heinz Heise Verlag.
IAH.
Lhomme, D., M. Dzikowski, G. Nicoud, B. Payraud, S. Fudral, and P.-L.
Guillot (1996). Les circulations actives des eaux souterraines des massifs
cristallins alpins: Exemple des Aiguilles Rouges (Haute-Savoie, France).
Compte rendu Academie des Sciences (serie IIa) 323, 681-688.
Lissey, A. (1968). Surficial Mapping 0/ Groundwater Flow Systems with Ap-
plication to the Oak River Basin, Manitoba. Ph. D. thesis, University of
Saskatchewan. 141 pp.
Lloyd, R. (1967). Oxygen-18 composition of oceanic sulfate. Science 156, 1228-
1231.
Lloyd, R. (1968). Oxygen-18 behaviour in the sulfate-water system. Journal 0/
Geophysical Research 73, 6099-6110.
Loew, S., B. Ehrminger, W. Klemenz, and D. Gilby (1996). Abschtzung
von Bergwasserzuflssen und Oberflchenauswirkungen am Beispiel des
Gotthard-Basistunnels. In B. Oddsson (Ed.), Instabile Hnge und andere
risikorelevante Natrliche Prozesse, pp. 353-376. Basel: Birkhuser Verlag.
Loew, S., H.-J. Ziegler, and F. Keller (2000). AlpTransit: Engineering geol-
ogy of the world's longest tunnel system. In GeoEng2000-an International
Con/erence on Geotechnical and Geological Engineering, Lancaster, Penn-
sylvania, pp. 927-937. Technomic Publishing Co.
Long, J., H. Endo, K. Karasaki, L. Pyrak, P. MacLean, and P. Witherspoon
(1985). Hydrologie behaviour of fracture networks. Mem. Int. Assoc. Hy-
114
drogeol. 17, 449-462.
Long, J., J. Remer, C. Wilson, and P. Witherspoon (1982). Porous media
equivalents for networks of discontinuous fractures. Water Resources Re-
search 18(3),645-658.
L6pez, D. L. and L. Smith (1995). Fluid flow in fault zones: Analysis of the
interplay of convective circulation and topographically driven groundwater
flow. Water Resources Research 31 (6), 1489-1503.
L6pez, D. L. and L. Smith (1996). Fluid flow in fault zones: Influence of
hydraulic anisotropy and heterogeneity on the fluid flow and heat transfer
regime. Water Resources Research 32(10),3227-3235.
Ltzenkirchen, V. and S. Loew (2001). Regional ground water flow in cata-
clastic fault zones. Results from large scale tracer test in the central Swiss
alps. In K.-P. Seiler (Ed.), Proceedings XXXI Congress IAH, Munieh. IAH.
submitted.
Mabee, S. B., K. C. Hardcastle, and D. U. Wise (1994). A method of collect-
ing and analyzing lineaments for regional-scale fractured- bedrock aquifer
studies. Ground Water 32(6), 884-894.
Malard, F., K. Tockner, and J. Ward (1999). Shifting dominance of sub-
catchment water sourees and flow paths in a glacial floodplain, Val Roseg,
Switzerland. Arctic, Antarctic and Alpine Research 31 (2), 135-150.
Maloszewski, P. and A. Zuber (1996). Manual on mathematical models in
isotope hydrology. Technical Report TECDOC-910, International Atomic
Energy Agency, Vienna.
Maloszewski, P. and A. Zuber (1998). A generallumped parameter model for
the interpretation of tracer data and transit time calculation in hydrologie
systems - comments. Journal of hydrology 204, 297-300.
Marechal, J.-C. (1998). Les circulations d'eau dans les massifs cristallins alpins
et leurs relations avec les ouvrages souterrains. Ph. D. thesis, Departement
de Genie Civil, GEOLEP, Ecole polytechnique federale de Lausanne, Lau-
sanne.
Martinec, J., H. Oeschger, U. Schotterer, and U. Siegenthaler (1982). Snowmelt
and groundwater storage in an alpine basin. In IAHS (Ed.), Hydrological
115
Aspects 0/ Alpine and High Mountain Areas, pp. 169-175. IARS.
Mazurek, M. (1998). Geology of the crystalline basement of northern switzer-
land and derivation of geological input data for safety assessment models.
Technical Report NTB93-12, NAGRA.
Menzies, J. (1995). Hydrology of glaciers. In J. Menzies (Ed.), Modern Glacial
Environments, pp. 197-240. Oxford: Butterworth-Heinemann.
Michelot, J. L. and J. C. Fontes (1987). Two case studies on the origin of aque-
ous sulphate in deep crystalline rocks. In Studies on Sulphur Isotope Vari-
ations in Nature, Vienna, pp. 65-75. International Atomic Energy Agency.
Monjoie, A. (1990). Impacts des travaux souterrains profonds sur les nappes
aquiferes en region montagneuse. In A. Parriaux (Ed.), Water Resources
in Mountainous Regions-Memoires 22nd Congres 0/ IAH, Lausanne, pp.
1208-1232. IAH: Ecole Polytechnique Federale de Lausanne.
Murdoch, L. C. and J. Franco (1994). The analysis of constant drawdown wells
using instantaneous source functions. Water Resources Research 30 (1),
117-124.
Nagra (1988). Berichterstattung ber die Untersuchungen der Phase I am po-
tentiellen Standort Piz Pian Grand (Gemeinden Mesocco und Rossa, GR).
Technical Report NTB88-19, NAGRA.
Neretnieks, I. (1983). A note on fracture flow mechanisms in the ground. Water
Resources Research 19, 364-370.
Neretnieks, I. (1993). Solute transport in fractured rock - applications to ra-
dionuc1ide waste repositories. In J. Bear, C.-F. Tsang, and G. de Marsily
(Eds.), Flow and Contaminant Transport in Fractured Rock, pp. 39-128.
San Diego: Academic Press.
NRC (1996). Rock Fractures and Fluid Flow. Washington D.C.: National Re-
search Council (NRC), National Academic Press.
Ofterdinger, U. S. (1999). Grundwassersysteme im geklfteten Kristallin des
westlichen Gotthard-massivs - Ttigkeitsbericht und Zwischenergebnisse.
Technical Report ETH3465/14, ETH-Zurich.
Ofterdinger, U. S. (2001). Regional Ground Water Flow Systems in the Ro-
tondo Granite, Central Alps (Switzerland). Ph. D. thesis, Federal Institute
116
of Technology (ETH), Zurich.
Ofterdinger, U. S., W. Balderer, S. Loew, and P. Renard (2000). Environmental
isotopes and hydrochemistry as indicators for ground water flow systems in
the rotondo granite. Ground Water. submitted.
Ofterdinger, U. S., P. Renard, and S. Loew (2001). Numerical modeling of the
ground water flow systems in the Rotondo granite. submitted.
Pearson, F., W. Balderer, H. Loosli, B. Lehmann, A. Matter, T. Peters,
H. Schmassmann, and A. Gautschi (1991). Applied isotope hydrogeology-
A case study in northern Switzerland. Technical Report 88-01, NAGRA.
Pearson, F. J. and C. T. Rightmire (1980). Sulphur and oxygen isotopes in
aqueous sulphur compounds. In P. Fritz and J. C. Fontes (Eds.), Handbook
of Environmental Isotope Geochemistry, pp. 227-258. Amsterdam: Elsevier
Scientific Publishing Company.
Pistre, S. (1993). Rle de la fracturation dans las circulations souterraines du
massif granitique de Millas (PyrEmees-Orientales). Compte rendu Academie
des Sciences (serie 11) 317, 1417-1424.
Priest, S. D. (1993). Discontinuity Analysis for Rock Engineering. London:
Chapman and Hall.
Pruess, K, B. Faybishenko, and G. Bodvarsson (1999). Alternative concepts
and approaches for modeling flow and transport in thick unsaturated zones
of fractured rocks. Journal of contaminant hydrology 38(1-3),281-322.
Rafter, T. and Y. Mizutani (1967). Oxygen isotopic composition of sulphates;
part 2, Prelimanary results on oxygen isotopic variation in sulphates and
relationship to their environment and to their delta (super34)S values. New
Zealand Journal of Science 10(3),816-840.
Rasmussen, T. and D. Evans (1993). Water infiltration into exposed fractured
rock surfaces. Soil Science Society Journal 57, 324-329. 2.
Raven, K (1977). Preliminary evaluation of structural and groundwater condi-
tions in underground mines and excavations. In Report of Activities, Num-
ber Paper 77-1a, pp. 39-42. Geological survey of Canada. Paper 77-1A.
Raven, K (1985). Field investigations of a small ground water flow system in
fractured monzonitic gneiss. In Hydrogeology of Rocks of Low Permeability-
117
Proceedings 17th International Congress IAH, pp. 72-86. Tucson, Arizona:
IAH.
Richards, K., M. Sharp, N. Arnold, A. Gurnell, M. Clark, M. Tranter,
P. Nienow, G. Brown, I. Willis, and W. Lawson (1996). An integrated ap-
proach to modelling hydrology and water quality in glacierized catchments.
Hydrological Processes 10,479-508.
Ritzi, R. W., S. Sorooshian, and P. A. Hsieh (1991). The estimation of fluid
flow properties from the response of water levels in wells to the combined
atmospheric and earth tide forces. Water Resources Research 27(5), 883-
893.
Rodhe, A. (1998). Snowmelt-dominated systems. In C. Kendall and J. J. Mc-
Donnell (Eds.), Isotope Tracers in Catchment Hydrology, pp. 391-433. Am-
sterdam: Elsevier.
Schlumberger (1985). Log interpretation charts. Technical report, Schlum-
berger Well Services.
Schlumberger (1989). Log interpretation. PrinciplesjApplications. Technical
Report SMP-7017, Schlumberger Educational Services, Texas.
Schneider, T. (1985a). Basistunnel-Furka, Los 63-Bedretto, Geologische Auf-
nahme des Fensters. Technical Report 338g, Furka-Oberalp-Bahn, Brig
(CH).
Schneider, T. (1985b). Geologischer Schlussbericht, Furka-Basistunnel. Tech-
nical Report 3380, Furka-Oberalp-Bahn AG, Brig (CH).
Schneider, T. R. (1976). Furka-Basistunnel- Zur Frage von Strungen im Bere-
ich des Fensterpunktes des Bahntunnels. Technical Report 338h, Furka-
Oberalp-Bahn.
Scholz, C. and M. Anders (1994). The permeability offaults. Technical Report
Open-File-Report 94-228, US geological survey. Proceedings of Workshop
LXIII - The mechanical involvement of fluids in faulting.
Schotterer, U., F. Oldfield, and K. Frhlich (1996). GNIP - Global Network for
Isotopes in Precipitation. Vienna: IAEA, WMO, IAHS.
Schotterer, U., T. Stocker, J. Hunziker, P. Buttet, and J.-P. Tripet (1995). Iso-
tope im Wasserkreislauf - Ein neues eidgenssisches Messnetz. GWA g, 1-8.
118
Sonderdruck Nr. 1348 des Schweizerischen Vereins des Gas- und Wasser-
faches (SVGW), Zrich.
Schwarcz, H. and G. Cortecci (1974). Isotopic analyses of spring and stream
water sulfate from the Italian Alps and Apennines. Ghemical Geology 13,
285-294.
Simmons, C., D. Hee Hong Wye, and P. Cook (1999). Modeling signal propa-
gation and weIl response in porous and fractured rock aquifers. In B. Fay-
bishenko (Ed.), Proc. Int. Symp. Dynamics of Fluids in Fraetured Rocks -
Goncepts and Recent Advances, Berkley (CA), pp. 79-83. Ernest Orlando
Lawrence Berkley National Laboratory.
Smith, L., T. Clemo, and M. D. Robertson (1990). New approaches to the
simulation of field-scale solute transport in fractured rocks. In S. Bachu
(Ed.), Proceedings Fifth Ganadian/American Conference on Hydrogeology,
Dublin, Ohio, pp. 329-342. National Water weIl association.
Smith, L., C. Forster, and J. Evans (1990). Interaction of faultzones, fluid flow
and heat transfer at the basin scale. In S. P. Neuman and I. Neretnieks
(Eds.), Hydrogeology of Low Permeability Environments, pp. 41-67. IAH.
Selected papers on hydrogeology, Volume 2.
Snow, D. T. (1968). Rock fracture spacings, openings, and porosities. Journal
of the soil mechanics and faundation division, ASCE 94 (SMI), 73-90.
Stehfest, H. (1970). Algorithm 368, numerical inversion of Laplace transform.
Commun. A GM 13, 624.
Stichler, W. and U. Schotterer (2000). From accumulation to discharge: Mod-
ification of stable isotopes during glacial and post-glacial processes. Hydro-
logical Processes 14, 1423-1438.
Taylor, B. E., M. C. Wheeler, and D. K. Nordstrom (1984). Stable isotope
geochemistry of acid mine drainage: Experimental oxidation of pyrite.
Geochimica et Gosmochimica Acta 48, 2669-2678.
Thury, M., A. Gautschi, M. Mazurek, W. Mller, H. Naef, F. Pearson,
S. Vomvoris, and W. Wilson (1994). Geology and hydrogeology of the
crystalline basement of northern Switzerland. Technical Report NTB93-01,
NAGRA.
119
Toran, L. and R. F. Harris (1989). Interpretation ofsulfur and oxygen isotopes
in biological and abiological sulfide oxidation. Geochimica et Cosmochimica
Acta 53, 2341-2348.
T6th, J. (1963). A theoretical analysis of groundwater flow in smaIl drainage
basins. Journal of Geophysical Research 68 (16), 4795-4812.
T6th, J. (1984). The role of regional gravity flow in the chemical and thermal
evolution of ground water. In B. Hitchon and W. EI. (Eds.), Practical Ap-
plications of Ground Water Geochemistry, Proc First Canadian/American
Conference on Hydrogeology, Worthington, Ohio, pp. 3-39. National water
weIl association and Alberta research counciI.
Tranter, M., G. H. Brown, A. J. Hodson, and A. M. GurneIl (1996). Hydro-
chemistry as an indicator of subglacial drainage system structure: A com-
parison of alpine ans sub-polar environments. Hydrological Processes 10,
541-556.
Van Everdingen, A. and W. Hurst (1953). The skin effect and its influence
on the productive capacity of the weIl. Trans. Am. Inst. Min. Metall. Pet.
Eng. 198,171-176.
Van Everdingen, R. and H. Krouse (1985). Isotope composition of sulphates
generated by bacterial and abiological oxidation. Nature 315, 395-396.
Velde, B. (1995). Composition and mineralogy of cay minerals. In B. Velde
(Ed.), Origin and Mineralogy of Clays, pp. 8-42. Berlin: Springer.
Vitvar, T. and J. Gurtz (1999). SpatiaIly distributed hydrologieal modeling
in the Rotondo area, western Gothard-massif. Technical Report 3465/16,
ETH-Zrich, Engineering Geology, ETH-Hnggerberg.
Vitvar, T. and U. Ofterdinger (2001). Hydrologie modelling of an alpine catch-
ment in the Gotthard-massif (central Switzerland). in prep.
Voborny, 0., P. Adank, W. Hrlimann, S. Vomvoris, and S. Mishra (1991).
Grimsel test site - modeling of groundwater flow in the rock body surround-
ing the underground laboratory. Technical Report NTB91-03, NAGRA.
Voborny, 0., S. Vomvoris, G. Wilson, G. ReseIe, and W. Hrlimann (1994). Hy-
drodynamic snthesis and modeling of groundwater flow in crystaline rocks
of northern switzerland. Technical Report NTB92-04, NAGRA.
120
Wallace, R. and H. Morris (1979). Characteristics of faults and shear zones as
seen in mines at depths as much as 2.5 km below the surface. Technical
Report Open File Report 79-1239, US geological Survey.
Wang, J. and T. Narasimhan (1993). Unsaturated flow in fractured porous
media. In J. Bear, C.-F. Tsang, and G. de Marsily (Eds.), Flow and Con-
taminant Transport in Fractured Rock, pp. 325-394. San Diego: Academic
Press.
Ward, J., F. Malard, K. Tockner, and U. Uehlinger (1999). Influence of ground
water on surface water conditions in a galcial flood plain of the swiss alps.
Hydrological Processes 13, 277-293.
Welch, P. (1967). The use of fast fourier transform for the estimation of power
spectra: A method based on time averaging over short, modified peri-
odograms. IEEE Trans. Audio Electroacoust. AU-15, 70-73.
Wenzel, H.-G. (1996). The nanogal software: Earth tide data processing pack-
age ETERNA 3.30. Bulletin d'information Marees terrestres 124, 9425-
9439.
Zaske, J. (1997). Einfluss eines Langzeitinjektionstests auf die Gezeitenant-
wort eines Bohrlochpegels am HDR-Standort Soultz-Sous-Forets. Master's
thesis, Geophysikalisches Institut Universitt Karlsruhe.
Zijl, W. (1999). Scale aspects of groundwater flow and transport systems. Hy-
drogeology Journal 7, 139-150.
Appendix A
Structural geology
121
122
The general geological setting of the study area has been outlined in paper 1
and 2. As earlier described, scanline surveys on the terrain surface as well as in
the Bedretto-tunnel provide the main database for the analysis of the structural
elements within the Rotondo granite, i.e. the orientation and frequency of the
observed discontinuities. In total, 23 scanlines were measured with lengths of
10-20 m. Scanline surveys were carried out in summer 1997 and 1998 as well
as in the Bedretto-tunnel in February 1998. Additional to these measurements
the record of the Borehole-televiewer-Iog, carried out in the research borehole
(Appendix B) was analyzed. The geological tunnel record of the Bedretto-tunnel
(Schneider 1985a) was used to extract furt her information on the orientation of
the prevailing fracturation within the Rotondo granite.
o
N
b)
o
N =531
a)
N
c)
Figure A.1: Stereographic projection (equal area) of the measured orientational
data; a) pole density plot of complete data set, b) of surface data set and c) of
tunnel data set
The measured data was corrected for sampling bias due to the scanline orien-
tation and individual measurements weighted accordingly for the stereographic
projection and cluster analysis (Priest 1993). From the above database four ma-
jor fracture sets were delineated (Table 3.1). Figure A.1 shows the stereographic
123
projection of the database (pole density distribution), distinguishing between sur-
face and tunnel data. Figure A.1c shows thereby that fracture sets 3 and 4 are
mainly associated to the data obtained from the tunnel surveys.
As a quantification of the spread within each set, the Fisher-distribution was
used to model the data of each set (Priest 1993; Fisher 1953). The magnitude
of the resuIting Fisher's constant, or precision, k (Table 3.1) yields thereby a
measure of the degree of c1ustering within each set. As the discontinuities ap-
proach parallelism, the Fisher's constant approaches infinity. In general, k ranges
between 5 and 30 for most field data (Priest 1993).
0.2
0.15
>-
tl
C
Q)
::l
0.1
er
&:
0.05
20 40 60 80 100 120 140 160 180
a) Azimuth [']
0.5
OA
>-
tl
c
0.3
Q)
::l
er
~ 0.2
LL
0.1
o
o 20 40 60 80 100 120 140 160 180
b) Azimuth [0]
Figure A.2: Azimuth-frequency histograms of measured fracture orientations in
the Bedretto-tunnel; a) histogram for complete tunnel data set and b) histogram
for discontinuities associated with tunnel infiows of ~ 2 l / m i n
For the calculation of the fracture frequencies of the individual sets, the ob-
served linear frequency along the sampling line A
s
(dependent on scanline orien-
tation) was corrected for the orientational bias to obtain the normal frequency
for each set A, i.e. the frequency along the normal to the set (independent of
scanline orientation; Priest (1993)).
With regard to the occurrence of ground water infiows to the gallery, the
azimuth-frequency histograms (Mabee et al. 1994) in Figure A.2 show that the
significant infiows ( ~ 2 Ijmin) are mainly associated to the fracture sets 1 and 2.
The measured data is stored on the attached CD-ROM (ascii-format). Figure
A.3 illustrates the organization of the data storage. The location of the scanlines
are referenced to the Swiss map coordinates in the case of the surface scanlines
124
CD-ROM
Istructural....Qoologyl
bedretto_tunnell surfacel
/
\ -----
scanlines_bedretto.dat tunneLrecord.dat bhtv.dat scanlines_rotondo.dat
Figure A.3: Organisation of data storage on the attached CD-ROM for individual
scanline data sets
or by the tunnel metrics in case of the scanlines measured in the Bedretto-tunnel
(for metrics refer to Figure 3.4).
Appendix B
Investigations related to the
research borehole
125
B.I
126
Borehole Ioeation and eonstruetion
In the scope of this study a research borehole was drilled from the subsurface gal-
leries within the Rotondo granite to obtain furt her information on the structural
geology of the granite and its hydraulic parameters. Hereby, a variety of methods
were applied ranging from geophysical borehole logging, hydraulic Packer testing
to presently ongoing long-term monitoring of hydraulic pressures.
o 25m
,'-------',
Figure B.1: Planview of the intersection of the Bedretto-tunnel and the Furka-
basetunnel (modified after Amberg Messtechnik AG (1994)). The location and
orientation of the research borehole are indicated.
The research borehole is situated at the intersection of the Bedretto- and
Furka-basetunnel (Figure B.1). It was drilled in March 1999 with a nominal di-
ameter of 115 mm to a length of 99.2 m. The borehole was advanced destructively
and indined with a flat 20 dip out of the horizontal and 30 from the axis of
the Bedretto-tunnel (Figure B.1). This specific orientation was chosen so that
the borehole would intersect the major prevailing fractures sets, which are ob-
served to be associated to tunnel inflows, at dose to right angles. During drilling,
several zones of water inflow to the borehole could be observed. Discharge from
the borehole increased sharply as the borehole advanced, e.g. increase from 1'V2
l/min for the top 44 m interval of the borehole to I'V15 l/min as the borehole was
advanced to 63 m. At the end of the drilling, the well-head discharge was in the
127
order of 20-23 l/min and after a few days stabilized at 22-23 l/min.
B.2 Geophysical borehole logging
After completion of the drilling a geophysicallogging campaign followed, applying
the following sondes:
- Borehole geometry sonde
- Natural-, sonde
- Natural-,-spectroscopy sonde
- Temperature/conductivity sonde
- Focused electric (guard log) sonde
- Impeller flowmeter
- Accoustic borehole televiewer
The equipment used was manufactured by Robertson Geologging, Deganwy
(UK) and provided by the Engineering Geology Institute.
With the borehole geometry sonde the borehole was first logged with regard
to the borehole azimuth, inclination and diameter (4-arm Caliper log). Borehole
wall break-outs or break-ins give first indications of 'weak' zones within the rock
mass.
The natural-,log is mainly used for depth correlation between individual logs,
as it is recorded together with each of the above listed logs. Similar to the natural-
, sonde, the natural-,- spectroscopy sonde measures the ,-radiation emitted by
the formation. This in turn is mainly associated to the radioactive decay of the
Uranium-Thorium-series, the Thorium-series and 4oK. The spectroscopy sonde
however enables adetermination and resolution of the individual energy speetra
of the measured ,-radiation and allows the estimation of the Uranium, Thorium
and Potassium content along the log. The concentration ratio of these elements
serves then as an indicator for the prevailing minerals in which theses elements
occur.
The focused electric sonde allows the determination of the formation resistiv-
ity of the rock mass by driving a focused alternating current into the formation.
128
Most formations will not conduct an electrical current, i.e. the rock matrix has
zero conductivity or infinite high resistivity. An electrical current will flow only
through the mineralized interstatial water within the formation. Thus minima in
the formation resistivity log may indicate a higher abundance of formation water
(Schlumberger 1989).
The temperature/conductivity sonde is used to measure the conductivity and
temperature of the borehole fluid. Varying gradients along the borehole indicate
potential inflows to the borehole.
The borehole televiewer is essentially an accoustic scanner, which scans the
borehole walls with a rotating transducer that emits a pulsed ultrasonic beam.
A visual representation of the pattern of accoustic reflectivity off the borehole
wall is displayed on a cathode-ray tube (Schlumberger 1989). On the basis of
this visual representation, discontinuities such as fractures and faults can be
identified and together with the orientational data from the borehole geometry
sonde, the orientation of these features can be determined (Appendix A).
The impeller flowmeter is applied to measure the flowrate in the borehole
fluid along the borehole. Changes in flowrate provide a direct measure of inflows
to the borehole. However, in boreholes with low inclination as in our case (20),
the applicability of the available impeller flowmeter is strongly limited and the
recorded data yielded no reliable results.
Figure B.3 illustrates the measured logs from the research borehole. In the
formation resistivity log it is apparent that the formation resistivity of the gran-
ite bedrock exceeds the measurement range of the sonde. For crystalline rocks
it has been shown that formation resistivity may exceed 60'000 stm (Hatzsch
1994). However, sharp minima in the resistivity log can still be observed, indi-
cating potential narrow zones of higher water content and potential inflow to the
borehole.
The caliper logs show the general decrease in borehole diameter as the bore-
hole was advanced. At a depth of rv20 m a sharp borehole wall break-out and
break-in can be observed in the X-axis and Y-axis respectively. This correlates
with the resistivity log, where a sharp minimum can be observed at this depth
interval. The borehole televiewer detects a distinct large fracture with an orien-
tation (120 strike) intermittent between fracture set 2 and 3 (Table 3.1) at this
129
depth. However, in the temperature/conductivity profile no c1ear indications for
inflows to the borehole can be observed at this depth.
In these latter profiles an anomalous minima can be observed in the temper-
ature log at a depth of rv30 m, correlated to a minimum in formation resistivity.
During the drilling of the borehole, inflows at this depth were observed. Thus
an inflow of slightly colder ground water at this depth may lead to the observed
temperature drop of the borehole fluid. The borehole televiewer identifies several
steeply dipping fractures of set 1 along this section. At the depth of rv55-65 m
strong variations in the temperature/conductivity profile are apparent, correlat-
ing with a pronounced variability of borehole wall break-outs in the Y-caliper
log as weIl as several sharp minima in the resistivity log. Observations during
the drilling also suggests major inflows to the borehole along this section. The
borehole televiewer detects numerous fractures of set 1 along this section. The
temperature/conductivity data implies that slightly warmer, more mineralized
water enters the borehole at this depth, off-setting the general slightly decreasing
trend in the temperature profile towards the borehole mouth .
To furt her investigate and quantify the inflows along the borehole a fluid log-
ging procedure was adopted and carried out. In this procedure, a fluid of known
conductivity (distinctly higher in contrast to the natural ground water inflows)
is injected into the borehole to replace the borehole fluid and under maintaining
a constant injection rate, temperature/conductivity profiles are logged along the
borehole. Natural inflows of lower conductivity lead to observable changes in
the conductivity profile of the borehole fluid. These can be used to detect and
quantify natural inflows to the borehole.
In our case, a NaCI-solution with a conductivity of 1880 J-lS cm-
1
was pro-
duced. The concentration was chosen to obtain a good contrast to the ground
water conductivity (rv 85-95 J-lS cm-
1
) while avoiding density effects. This so-
lution was injected at the borehole bottom until no furt her observable changes
of borehole fluid conductivity in the well-head discharge was measured. Follow-
ing the replacement of the borehole fluid temperature/conductivity profiles were
logged down- and uphole, maintaining the constant injection rate of 20 l/min of
the NaCI-solution at the weIl bottom. Figure B.4 shows the results of this test.
Several sharp changes in fluid conductivity can be observed, especially in the
(B.1)
130
downhole-logs (Iogging-run against the flow in the weIl), which better resolves
the low conductivity inflows. For an estimation of the inflow rate in individual
depth intervals, a mixing ca1culation was applied starting from the injection point
at the bottom of the borehole (equation B.1).
Clb X qlb +Ci X qi
Cub =
qlb + qi
From the fluid 10gging the known flow rate (qlb) and fluid conductivity (clb)at
the lower bound of an investigated depth interval and the known borehole fluid
conductivity at the upper bound of the investigated interval (Cub) are extracted.
Together with the known conductivity of the inflowing ground water (Ci), ex-
tracted from the undisturbed conductivity log (Figure BAd), the unknown inflow
rate (qi) was ca1culated. For the latter the assumption was made that the mea-
sured fluid conductivity in the undisturbed log can be used as estimate for the
conductivity of the inflowing ground water. This is however not exactly valid, as
the measured fluid conductivity in the undisturbed conductivity log also already
represents some degree of mixing of inflows and borehole fluid under undisturbed
conditions. Yet, as the variations and absolute values of the conductivity in
the undisturbed log are much smaller than in the conductvity log under test
conditions, the applied simplification is feasible to approximate inflow rates for
individual depth intervals. The potential error introduced by this simplification
would be more significant if inflows of high contrasting conductvity were expected
along the borehole. Field measurements in the gallery however show that such
inflows of high contrasting conductivity can not be observed in the Rotondo gran-
ite (Paper 1, Appendix C). The approximated inflow rates are stated in Figure
BAe. They reproduce weIl the drilling observations of water occurrence (Figure
B.3h) and correlate also weIl with the observations from the other logs with re-
gard to potential inflows to the borehole (Figure B.3b-f). The ca1culated total
discharge from the borehole amounts to 24.8 Ijmin, which is in agreement with
the measured discharge rate at the weIl-head of 22-23ljmin.
As mentioned above, the diagnostic plot of the Thorium/Potassium ratio gives
an indication of the prevailing clay mineralogy along the borehole. The diagnostic
plot of the Th/K-ratio (Figure B.2b) reveals a large scatter of the data within the
field ofmixed layer clays (e.g. Illite/Smectite, Chlorite/Smectite (Velde 1995)), Il-
131
4
cK(%)
30
a)
_.J'r}if'J;.5
----
30
40
t!lJ <V
jJ Ar
i I T
iIll /
/
t/ //
Ipo\W't--
e
/
I //
i5 I /
5:-1 / _-
JI / _---
10 __ - \\\\\.e _---
/ _ _ - . Feldspars
J(j / ..f>{;;' - - - - - . Glau90nltel:

o e.- - :;, _ - 51 -. - - ... ""'" Potassiu,m Evaporihis
o 1 4
cK(%)
E
"-
.s 20
.<::
to
3 4
cK(%)
Figure B.2: Diagnostic plot (after Schlumberger (1985)) of Potassium and Tho-
rium concentrations from the ,-spectroscopy data. Figure B.2a illustrates the
correlation between the Th/K-ratio and c1ay mineralogy. Figure B.2b shows the
diagnostic plot of the complete data set, c) a plot for the data set correlating to
anomalies in the temperature log, d) subset correlating to natural-, anomalies
and e) correlating to anomalies in the formation resistivity log.
lite, Micas and partly Chlorite. To investigate whether anomalies in the profile of
other logged parameters, which in turn might be attributed to denser fracturation
and/or discrete inflows to the borehole, are correlated to a specific c1ay mineral-
ogy, i.e. preferential coating of fracture planes by specific c1ay minerals, the total
depth record was differentiated into several subsets. These subsets correspond
to depth intervals where anomalies in the temperature, natural-, or formation
resistivity log were observed (Figure B.2c-e). No c1ear correlation of parameter
anomalies with regard to the c1ay mineralogy can be observed. Only in the subset
of natural-, and formation resistivity anomalies it appears that a higher fraction
of the ,-spectroscopy-data indicates the presence of Illite. However, the scatter
of the data is still substantial. Striking is the lack of detected Chlorite in the
,-spectroscopy log. Observations in the Bedretto-tunnel show that many frac-
tures are characterized by Chlorite-coatings. In a comprehensive study within
the similar crystalline basement rocks of northern Switzerland, Mazurek (1998)
reports that the c1ay fraction of both altered whole-rock and fracture coatings
132
are mainly dominated by Illite, mixed-layer c1ays and Chlorite. The individual
fractions of the specific minerals showed thereby strang local variability. The
Th/K-ratios with the indicated lack of Chlorite along the borehole could thus be
explained by local variability at this specific location within the granite massif
when compared to the observation along other sections of the Bedretto-tunnel.
Metresl Natural ... Calil,erX\x;< ...... 1.
o (cps) 2501.)(} Im",) Inun) 140l
"
11
U
rluidTemp
iO 121 IT,
Fluid Cond. I roml Res.
22180 100 5110 mmj J4i101l
..
whilc drilling
progress ('week

h)
",ogre..
:)

%.-1.0'"
-
water
...
i;c,l) 1<\.l;)rrJ
'weak rock' watet

end oflool
(ppm) )l}
nominal diameter
-80
-60
-40
-20
Figurc B.3: Rcsults of thc gcophysical borcholc logging; a) natural-'Y log, b) calipcr log (X-axis), c) calipcr log (Y-axis), cl)
'Y-spectroscopy log, e) fluid temperature log, f) fluid conductivity log and g) formation resistivity log. Figure B.3h states
the observations made during the drilling.
Metres I I?9W!'f=l".L I
lImiD
27.S0JfI.SOm: I.n llmin
30.50 _ SO.4<im : 0_07 I/min
50.46 51.63m; 2.3 Vrnm
lIrniD
55.14 5tl.JMm : 4.39 lImin
SS.JM S'J.J5m: .1.19 IImin
59.35 61.7-4m: 0.16 !lmjn
61.74 M.l6Dl: 5,04 lImm
65.16. 6S.9Om: 0.04 Jlmln
(,5.')() (r).')Km : O.lms l/min
69.9R 1IU4m: 0.12 I/min
76.14 - 85.49ul: 0.4 lImin
'15.49 _ tloll.62nl : 0.31 11min
XKf>2 ')].f>Um : 11.1)5 lImm
93.60 _ 99.l0m : 4.0 Jlmln
O.961'min
1,ol/min
O.07!'min
calculated Inflow
Tbc calculutcd influws aOlouaf to 24.g Vmin in
total. This ilJ in good agreement wlth dililChnrge
1
Fluid Cond.
'.;I.uiAT\;:-'11r
J
tel n() 1(",.;1
(J,tS/Cnl) no test
2.3Vmln
2.1\ IImin
4.391
'
min
3_.19.I'min
(j,161'mln
5.04l'mlu
./ .a:o.lI'mili
0,005 Umin
O.12l'min
UAlIuilil
lim,"
;1
\ r-
o.os Fmin
4.01/1'1'111
e)
UP FL2 Fluid Cond.
(I-'S/cm) 2000 Iso
DO_"T1_FIII,d ('""d.
.- ... ?OflO 1I <)
22 I i-:O
l)q
300119 ce,
l-- UP FL2 Fluid Temp.
19 ("C)
o
-80
-60
-40
-20
-100
Figure BA: Results from the fluid-Iogging test; a)natural-J' log, b) fluid temperature log, c) fluid conductivity log, d)
fluid temperature and conductivity log under undisturbed conditions and e) calculated inflow rates to the borehole for
individual depth intervals (in BAb and BAt: DO= down-hole logging run, UP = up-hole logging run).
B.3
135
Hydraulic Packer-testing
Following the geophysical investigations in the research borehole, hydraulic packer
tests were carried out to determine hydraulic parameters for individual depth in-
tervals. The testing equipment was manufactured by Solexperts, Schwerzenbach
(eH) and provided by the Engineering Geology Institute. Testing was targeted
to investigate those depth intervals showing ground water inflows to the borehole
as detected by the geophysical logs and to obtain first estimates of hydraulic
conductivities for the moderately fractured granite. However, first double packer
tests, separating single borehole intervals for testing, revealed that after interval
shut-in build-up pressures rapidly exceeded the manageable working pressure of
the testing equipment (35 bars). For this reason, double packer tests of individ-
ual depth intervals were not possible. As a modification, single packer tests were
performed for increasing borehole sections, starting at the weIl bottom.
The tests were carried out by reducing the artesian discharge from the in-
vestigated section at the weIl-head, maintaining a constant discharge rate during
individual tests. The succeeding pressure build-up in the packed-off borehole sec-
tion was recorded downhole by piezorestive pressure transducers. For the data
analysis the tests were superimposed as constant-rate injection tests. Due to
technical difficulties only four tests could be performed with test durations of 6-9
hours. Tested borehole sections were thereby:
- 99.2-93.0 m (section 1)
- 99.2-80.4 m (section 2)
- 99.2-65.6 m (section 3)
- 99.2-63.5 m (section 4)
The observed pressure response in the individual tests varied from a behaviour
usually encountered in isotropic homogeneous reservoirs (section 1, Figure B.5)
to double porosity behaviour as observed in section 4 (Figure B.6).
The pressure response in sections 2 and 3 showed a peculiar behaviour with a
steadily increasing derivative a late times of slope a = 0.22 and 0.29 respectively.
Several reservoir models (homogeneous isotropic and double porosity (Bourdet
136
10
3
:r- --,
1
1
1
1
I
I
I
1 1

1 1
1 I
I 1
I 1
1
10
2
- - - - - - - - - -
1
1
1
1
1
-----------,-----------
1
: 1" dP
1 D dP'
1
1
I
1
15=3.2
o I C =1.5x 10-
6
m
3
/kPa
10 -/--,--,--,-,--rrr-rr---.----r--,-r-r1-rr!----.----,-,--.-r---rn+--,--,-r-r",r-rr-"---'-T-'-'-,.:-r'-i'--Mr'ri-i
10
0
dt [sec]
Figure B.5: Analysis of packer test in section 1 (99.2-93m). Match of the mea-
sured data (pressure response and derivative) with homogeneous isotropie reser-
voir model. Indicated are skin factor Sand wellbore storage C.
10
3
,-------------,--------,-------,---------,
11..
"Cl
06
a.
"Cl
I
I
1
I
1
I
1
I
1
1
1
- - - - - - - - - - -1- -
1
1
1
1
I
1
I
I
1 I
---------T-----------r-----------
1 I a
1 I
I
1
+ dP
D dP'
5 =0.8
C= 1.6x 10-
6
m
3
/kPa
w=2.9x10-
3
10

dt [sec]
Figure B.6: Analysis of packer test in section 4 (99.2-63.5 m). Match of the
measured data (pressure response and derivative) with double porosity model
(Bourdet and Gringarten 1980). Indicated are skin factor S, wellbore storage C,
storativity ratio wand interporosity flow coefficient A.
137
10
3
:r------c-------c--------,--------,---------,
.-'
... -----
1
----- ... --------
~ . ,
: ,,/
: / ...
- - - - - - - - - - - - - - : , ~ - - - -
A ~
/",0.
~
./' .
,
/"
#
-f"-,- - -- - - - ---------. - - - - - - -- -- - - ---
10
2
(il
a.
6
CL
""Cl
0/5
a.
""Cl
10
'
:2
:S =35
:C = 1:3x10-
6
m
3
/kPa
:1 :L
,
=42m
:S=2 :L
2
= 16 m
:C= 1.3x 10-
6
m
3
/kPa:y = 18
10 +----,----,---r-rTTTr+------,----,--.--rTTTrr_----,----,---r-r"TT"T-r+------,----,----,-,.....,.rrt-'--,---,---.-r-rTrn
10 10
'
10
2
10
3
10
4
10
5
dt [sec]
Figure B.7: Example for packer test result with steadily increasing late time
pressure derivative (borehole section 3, 99.2-65.6 m) Match of the measured data
with homogeneous isotropie reservoir model; 1) infinite reservoir, 2) reservoir
bound in two directions by intersecting sealing faults in distance L1 and L2 from
the weIl. Indicated are skin factor S, weIlbore storage C as weIl as distance to
faults (L1 and L2) and intersection angle of the faults I for model 2.
and Gringarten 1980)), weIl models (with storage & skin or with a vertical frac-
ture (Bourdet and Gringarten 1980; Gringarten et al. 1974; Cinco-Ley and Meng
1988)) as weIl as different boundary scenarios were applied in the attempt to
reproduce this behaviour, yet without satisfactory result. A more complex geom-
etry of intersecting sealing faults lead thereby to a doser but still unsatisfactory
approximation of late time derivative behaviour (Figure B. 7). Furthermore, ge-
ological observations in the borehole and along the gaIleries do not indicate the
presence of sealing faults in the vicinity of the borehole and therefore the approach
of complex fault geometries appears less feasible.
One furt her possible explanation for the steady increase of the late time deriva-
tive is provided by the generalized radial flow model (GRF; Barker (1988)) and
the furt her development of this model by Hamm and Bidaux (1996). In this
context, the steadily increasing late time derivative data would correspond to
a non-integral flow dimension of n=(2-2a) = 1.56 and 1.42 respectively (Figure
8.8 and B.9). The high skin factor necessary in the match of the model to the
138
I
Ql
'"
C
00
o
o
o
o
+ watertevel
o derivative
n =1.42
kf b3-n =1.8 x1 0.
6
m4-n/s
C =1.1x10-
5
m
2
S = 120
100
10
1
10
2
10
3
10
4
10
5
elapsed time (sec)
pressure derivative (borehole section 3, 99.2-65.6 m) Match of the measured data
with generalized radial flow model (Barker (1988), (Hamm and Bidaux 1996)).
Indicated are flow dimension n, generalized transmissivity kfb
3
-
n
(b extend of
flow region), weIlbore storage C and skin factor S.
data could be explained according to Hamm and Bidaux (1996) as a geomet-
rie effect ('completion skin', similar to partial penetration) related to the shape
of the weIl in the n-dimensional space, rather than the traditional skin damage
(Van Everdingen and Hurst 1953). The interpretation of this non-integral flow
dimension and the generalized transmissivity (kfb
3
-
n
) is however not straight-
forward. Barker (1988) suggests that the dimension is related to a characteristic
property, e.g. the fractal dimension (Long et al. 1985) of the fracture network
and a specific topology of the flow system. The non-integral flow dimension is
thereby not an intrinsic hydraulic property of the fracture system (Barker 1988),
which might explain why some of the tests reproduce this behaviour while others
do not. Table B.1 gives an overview of the deduced transmissivities and apparent
hydraulic conductivities (referenced to section length) for the individual tests.
The test data is stored on the attached CD-ROM (ascii-format). Figure B.10
shows the organization of the data storage.
139
80
n ~ 1.42
70
kf b3-n = 1.8 x1 0-
6
m
4
-
n
/s
C =1.1x10-
5
m
2
S ~ 120
60 o waterlevel
o derivative
I
50
Q)
'"
C
Q) 40
>
Q)
~
CU 30
;:
20
10
0
10
1
00 00 00
elapsed time (sec)
pressure derivative (borehole section 3, 9 9 . 2 ~ 6 5 . 6 m) Match of the measured data
with generalized radial f1.ow model in semi-logarithmic plot (Barker 1988; Hamm
and Bidaux 1996). Indicated are f1.ow dimension n, generalized transmissivity
k
f
b
3
-
n
(b extend of f1.ow region), wellbore storage C and skin factor S.
Table B.1: Transmissivities and apparent hydraulic conductivities (referenced to
section length) deduced from packer tests of the individual borehole sections. (t
value derived from first infinite acting radial f1.ow period)
Tested section of the
borehole
99.2-93.0 m
99.2-80.4 m
99.2-65.6 m
99.2-63.5 m
Hydraulic conductivity
(m/s)
2.9xlO-
8
1.2x 10-
8
t
6.0x10-
9
5.0 X10-
9
t
Transmissivity
(m
2
/s)
1.8x10-
7
2.2x10-
7t
2.0x10-
7
1.8x10-
7t
140
CD-ROM
~ a c k e r t e s t ~
packertest_section1.dat packertest_section2.dat packertesCsection3.dat packertest_section4.dat
Figure B.10: Organisation of data storage of individual packer tests on the at-
tached CD-ROM.
B.4 Borehole completion and long-term mon-
itoring of hydraulic pressures in borehole
intervals
BA.1 Borehole completion
Once packer testing was completed the borehole was equipped for the long-term
monitoring of hydraulic pressures in two separated borehole intervals. For the
monitoring the interval between 50 m and 66 m, where considerable inflows were
detected in the geophysicallogs and the bottom interval ofthe borehole (2:::93m),
farthest away from the gallery were chosen. The intervals were isolated from
the borehole sections above and below by grout injections (cement packers) and
the remaining borehole top section was additionally filled with grout. Hydraulic
pressures are recorded at the borehole mouth via I" tubings to the intervals
(Paper 2). Figure B.11 illustrates the borehole completion.
."
CD'
'"'
s-
g
~ .
:::
'
0 0 0
Q
0 0 0 0
g. CD
W
CD CD CD CD CD CD CD
s-
::: ::: ::: ::: ::: ::: ::: ::: :::
~ . '
~ ..... ~ ~ ~ ~ ~ ~ ~
() CD
~ ~ ~ ~ ~ ~ ~ ~ g.
'"
CD CD CD CD CD CD CD CD
::: .... .... .... .... .... .... .... ....
' Co 'l O'J <J1 -l:>. W
'"
CD
10m 25m 40m 50m 56m 66m 76m 86m
I
I-'
,+:..
I-'
Pressure
gauge 2 gauge 1
m
Cement packer 2
~ I " ,1/
16m
Monitoring /nterva/ 2
m
Cement packer 1
/1' "I
13.2m
Monitoring /nterva/1
Figure B.ll: Schematic sketch f borehle cmpletin fr lng-term mnitring f hydraulic pressures
142
52
50
48
ffi
oe.
~
iil 46
(/)
~
9-
m
C: 44
Q)
:5
42
40
38
1.10.1999 1.01.2000 1.04.2000
Date
1.07.2000 1.10.2000
Figure B.12: Long-term record of hydraulic pressures in the research borehole;
data from monitoring interval 1 and 2.
B.4.2 Long-term hydraulic pressure record
Monitoring of hydraulic pressures in the isolated borehole intervals (Figure B.11)
commenced in July 1999 and is presently ongoing. After a preliminary monitor-
ing phase the final sensor setup with piezoresistive transmitters was installed in
October 1999 (Type PA-53, Keller, CH). Hydraulic pressures were recorded on
a 10 minutes sampling interval. Figure B.12 illustrates the recorded hydraulic
pressures. The recorded signal in interval 1 shows an annual fluctuation in the
order of 1-1. 5 bars with maximum pressure in November and minimum pressure
in June. Whereas pressures in interval 2 show a similar dedining response during
winter/spring 1999/2000, the pressure rise towards autumn 2000 is less dear.
Comparing the measured pressure signals with the typical annual ground wa-
ter recharge pattern in the upper Gerental valley, extracted as long-term monthly
average from the results of the hydrological model (9-year average), a time lag be-
tween maximum recharge rates and maximum hydraulic pressures of 3-4 months
can be observed (Figure B.13). This time lag is most probably governed by a
143
0.35
0.30
<J)
~
CU 0.25
J::
(,)
~
ro
0.20
:::J
c:
c:
CU
0.15
'0
c:
0
0.10
t5
~
005
000 + - - ~ - _ - - - . - . = ......
Jan Feb Mar Apr May Jun Jul Aug Sep Oel Nov Dec
Date
Figure B.13: Long-term average ground water recharge pattern in the upper
Gerental-vaIley.
combination of the saturated hydraulic diffusivity of the granite massif and the
delay caused by the transition of the ground water recharge through the un-
saturated zone before reaching the free ground water table. The thickness of
the unsaturated zone has been shown to be significant (Paper 2). Conceptual
models on the flow through the unsaturated zone of fractured rocks differ, at-
tributing downward infiltration to either saturated matrix blocks (e.g. Wang and
Narasimhan (1993)) or fractures (e.g. Pruess et al. (1999)). In a field study
where both pathways were identified, recharge pulses (in this case stream seep-
age) was observed to penetrate to a depth of 150 m in a matter of days to weeks
(NRC 1996). Another field study showed a time lag between recharge pulse (pre-
cipitation) and ground water inflow variations into a 120 m deep mine of only
several days (Bassett (1997) as quoted by Pruess et al. (1999)). Considering the
simulated thickness of the unsaturated zone in the Rotondo area, a delay of the
penetration of ground water recharge to the water table in the order of several
weeks could weIl be contributing to the observed time-lag between peak recharge
rates and measured maxima in hydraulic pressure in the borehole.
The recorded pressure data is stored on the attached CD-ROM (Figure B.14).
144
CD-ROM
Ilongterm_monitoringl
p_intervaI1.dat
-------
p_intervaI2.dat
Figure B.14: Organisation of data storage of long-term pressure reeord of interval
1 and 2 on the attaehed CD-ROM.
BA.3 Effects of tidal forcing
To investigate potential furt her influenees affecting the natural behaviour of the
ground water system the reeorded hydraulic pressure signal was also analyzed
with regard to temporal small seale fluetuations. In partieular we were interested
to see whether earth tide effeets and influenees of barometrie pressure variations
(meteorologieal variations and atmospheric tides) are refleeted in the reeorded
hydraulie pressures in the borehole.
AprOO
cl
MarOO FebOO JanOO
Dec99
Dec99
0.04
~
0.02
~
0
n.
"0
-0.02
-0.04
Nov99
0.04
~ 0.02
~
n. 0
"0
-0.02
-0.04
Nov99
0.84
~
Ol
0.82
a
l!!
0.8 ::>
VI
VI
~
0.78
n.
0.76
Nov99
Date
Figure B.15: Residual hydraulic pressure after removal of the linear deereasing
trend in the original pressure signal (see Figure B.12) in a) interval 1 and b)
interval 2; e) barometrie pressure measured at the meteorologieal station Grim-
seI/Hospiz. Indieated is the date of the 'Lothar storm event'.
145
The analyzed data spans the period from 14.11.1999 to 31.03.2000. Hydraulic
pressure data was taken from the borehole intervals 1 and 2 (Figure B.11). Baro-
metrie pressures were taken from the nearest meteorologieal station at Grim-
seI/Hospiz (1980 m a.s.l), approximately 9 km northwest of the borehole loeation.
8ampling interval for the datasets was 10 minutes.
Preliminary plots (full period, weekly and daily close-ups) of the pressure
signals (Interval 1,2 and barometrie) in the time-domain reveal that direet influ-
enees from barometrie pressure ehanges are not clearly apparent. The amplitudes
of the meteorologieal indueed barometrie pressure ehanges are generally smaller
than the fluetuations in the interval pressures (Figure B.15). 8peeifie major
events like the 'Lothar storm event' (late Deeember) are not clearly reprodueed
in the pressure signal of interval 1 and 2.
X10-4
6
8, K,
Inlerval 1
'6'
0.
5 -
et
'"
M2
g
~
4 -
E
~
Q)
.i':'
3 -
'in
c
Q)
"C
l" 2 0, -
tJ
~
"'
W
1
~ 8 ~
-
~
Cl.
0
MAA AM
'N
J\
A.
0.5 1.5 2 2.5
Frequency [cpdl
Figure B.16: Power speetrum density plot for hydraulic pressures reeorded in
interval 1 (epd: eycles per day). Indieated are the earth and atmospherie tides
associated to the individual peaks in the speetrum.
Further investigations were earried out in the frequeney domain. The original
interval signals were first filtered with a 4
th
order Butterworth highpass filter at
the cut-off frequeney of 0.06 epd (removal of large seale trends; cpd eycles per
day). Then using the Welch-method (Welch 1967) the power speetral densities
were eomputed (Figure B.16 and B.17). The power speetra of interval 1 and 2
show eaeh clearly four peaks. These ean be attributed to the main lunar tides
0
1
and M
2
and the atmospherie tide 82 at diurnal and semi-diurnal frequeneies
respeetively (Zaske 1997). As the frequencies for the diurnal atmospherie tide
146
SI and the diurnallunisolar tide K
1
oeeur at similar frequeneies a distinetion on
the basis of the available data is not possible. With regard to Figure B.16, B.17
and B.15 we ean state that the influenee of the tidal foreing funetion, i.e. the
earth tide dilation acting on the granite matrix and in partieular on the subver-
tieal fractures, is clearly deteeted in the pressure reeord of both intervals. Small
seale meteorologieal indueed variations in barometrie pressure are not reprodueed
by the pressure reeord in the borehole. However, the effeet of the periodie at-
mospherie tides (barometrie loading), mainly eaused by thermal effects of solar
origin, is also reeorded on a diurnal and semi-diurnal frequeney. With regard to
the subvertieal orientation offraetures and faults within the granite roekmass, we
ean assume that hereby the reeorded signal is mainly attributed to the horizontal
areal strain due to the flexur of the surfaee region as response to the barometrie
loading.
X 10-
4
Interval2
1.5
Frequency [cpd]
2.5
Figure B.17: Power speetrum density plot for hydraulic pressures reeorded in
interval 2 (epd: eycles per day). Indieated are the earth and atmospherie tides
assoeiated to the individual peaks in the speetrum.
Comparing the power speetrum density plots of the two intervals, we ean ob-
serve that deteeted signal amplitudes are generally larger in interval 1. Simmons
et al. (1999) report in their study on signal propagation that higher storativities
enhanee signal attenuation. They furt her show that in fraetured environments,
signal amplitudes deerease linearly with inereasing fraeture aperture width. The
geophysieal logs indieate denser fraeturation in interval 2 and thus a larger total
aperture width as in interval 1 ean be assumed.
147
The original pressure signals of interval 1 and 2 were also filtered in the time
domain to retrieve the residual pressure signal in the range of the observed tidal
frequencies (!"VI and!"V2 cpd). Hereby, a high-pass 4
th
order Butterworth filter with
a cut-off frequency of 0.5 cpd and a low-pass filter with a cut-off frequency of 1.5
cpd were applied to retrieve the remaining original signal at the diurnal frequency.
Likewise 4
th
order Butterworth filters with cut-off frequencies of 1.3 and 2.5 cpd
were used to compare the remaining signal for the semi-diurnal frequency. In
Figure B.18 and B.19 the residual signal for interval 1 is examplarily illustrated
in the time domain far the investigated measurement period.
0.015r----.....------,---.------.-------,-----,-------,
-0.015 '---__-'-__---'- .L.-__-"-__----' -'--__-'
o 20 40 60 80 100 120
TIme [d]
Figure B.18: Residual pressure variations in interval 1 in the diurnal frequency
range.
0.015r----.....------,---.------.-------,-----,-------,
-0.015'------'------'----.L.-----l.-------'----'-----'
o 20 40 60 80 100 120
Time [d]
Figure B.19: Residual pressure variations in interval 1 in the semi-diurnal fre-
quency range.
Signal amplitudes lie hereby in the range of 1-10 mbar. At the Grimsel test
148
site in the northern vicinity of the study area, earth tide effects up to 15 mbar
are reported (Frick et al. 1992).
With regard to any furt her detailed interpretations of the observed power
spectra, e.g. as to why the SlKl-amplitude in interval 1 is larger than in interval
2, one fundamental restrietion needs to be made. The sensor-setup of the moni-
toring system was not chosen to investigate specifically the above phenomena. So
when analyzing the available data for these small scale pressure variations, the
transducers are working in fact at the lower end of their measurement resolution
as specified by the manufacturer. Therefore interpretations should be restricted
to the above basic observations and care should be taken not to produce too
elaborate interpretations based on the available data set. However, if sensors of
higher precision were to be used for future pressure monitoring and tidal forcing
functions were approximated by synthetical calculations (Wenzel 1996), the anal-
ysis of the pressure response to tidal and barometrie forcing appears a promising
tool at this site to investigate and quantify in situ poroelastic parameters of the
granite (e.g. shear modulus, uniaxialloading efficiency and Biot 's constant) and
valuable hydrogeological parameters such as specific storage. The observed data
suggests that therefore the analysis of the response at the frequencies of the main
lunar tides 0
1
and M
2
is the most promising, as they can be unambiguously
resolved. Studies of Beavan et al. (1991), Evans et al. (1991) and Ritzi et al.
(1991) have demonstrated for example the feasibility of this approach for the
determination of the above parameters.
Appendix C
Hydrochemistry
149
150
SampIes of the encountered ground water were collected from spring dis-
charges at the terrain surface in the Rotondo area as well as from inflows to
the Bedretto-tunnel. In the following, measured in situ parameters and results
of the hydrochemical analysis of collected ground water sampIes are tabulated
(Tables C.3-C.14). The major ionic composition was analyzed by ion-exchange
liquid chromatography while parameters such as pH, temperature, conductivity,
alkalinity and oxygen content were measured in situ during sampling with elec-
tronic probes or field titration respectively. Table C.l gives the location of the
individual sampling sites at the terrain surface. SampIes from the Bedretto-tunnel
are referenced to the tunnel metrics (Table C.2, Figure 3.4).
151
Table C.1: Location of the sampling sites at the terrain surface. Listed are the
8wiss map coordinates of the sampling locations as weIl as the geology at the
individual site (RG Rotondo granite; 0 orthogneiss; refer to Figure 2.1).
8ampling location Geology X-coordinate Y-coordinate Altitude
[m] [m] [m a.s.l.]
ch1 RG 676370 151400 2240
ch2 RG 676260 151430 2240
hw RG 676340 151880 2200
met RG 675310 151060 2300
swf 0 674370 152670 1820
Table C.2: Location of the sampling sites in the Bedretto-tunnel referenced to
the tunnel metrics (Figure 3.4). All sampling locations are situated within the
Rotondo granite.
8ampling location
Tl
T2
T3
T4
T5
T6
T7
T9
Tll
81
82
Tunnel metrics TM [m]
5074
5000
4945
4818
4758
4658
4490
4208
4464
4272
3890
Table C.3: In situ parameters and chemical analysis of ground water sampIes collected in the Rotondo area (Temp. sampIe
temperature, Cond. electrical conductivity, T DS total dissolved solids, nn not detected).
Parameter SampIe
ch1 ch1 ch1 ch2 ch2 ch2 hw hw hw met swf swf swf
Date 1.10.97 1.9.98 12.8.99 1.10.97 2.9.98 12.8.99 1.10.97 2.9.98 15.8.99 31.8.99 5.10.97 2.9.98 5.8.99
Temp. [aC] 8.9 7.5 8.6 8.5 8.9 9.5 6 4.2 7.5 9.5 6.2 7.3 6
Cond. 14 18 15 16 19.5 15 15.2 15.5 15.8 15 64 68 62
pH 7.4 7.7 6.8 6.9 7.3 7.2 6.8 7 7.4 7.5 7.1 7.2 7.7
O
2
[mg/l] 7.7 8.3 7.9 8 7.7 7.2 8.2 10.2 7.5 8.3 6.8 6.5 6.9
TD8 [mg/ll 11.8 8.8 11.9 12.7 12.8 12.5 14.4 15.3 14.9 8.8 46.5 46.6 46.3
Li [mg/ll nn nn nn nn nn nn nn nn nn nn nn nn nn
Na [mg/l] 0.8 0.77 0.81 0.85 0.91 0.8 1.8 1.91 1.85 0.8 0.84 0.88 0.8
I
K [mg/l] 0.42 0.44 0.41 0.4 0.4 0.44 0.76 0.89 0.78 0.41 1.54 1.58 1.62
Mg [mg/l] 0.12 0.04 0.14 0.1 0.12 0.06 0.1 0.14 0.12 0.06 0.82 0.76 0.81
Ca [mg/l] 2.05 1.23 2.03 2.6 2.65 2.55 2.16 2.21 2.2 1.4 9.2 9.3 9.2
NH4 [mg/l] nn 0.25 nn nn nn nn nn nn nn nn nn nn nn
F [mg/l] 0.22 0.55 0.21 0.36 0.34 0.3 0.28 0.25 0.28 0.6 0.14 0.12 0.1
Cl [mg/l] 0.21 0.16 0.44 0.22 0.19 0.15 1 1.55 1.1 0.12 0.17 0.21 0.15
Br [mg/l] nn nn nn nn nn nn nn nn nn nn nn nn nn
80
4
[mg/l] 1.61 1.62 1.58 2.1 2.06 2.12 1.62 1.57 2 1.7 3.3 3.21 3.1
N0
3
[mg/l] 0.22 0.64 0.23 nn nn nn 0.6 0.69 0.42 0.6 nn nn nn
N0
2
[mg/l] nn nn nn nn nn nn nn nn nn nn nn nn nn
HC0
3
[mg/l] 6.1 3.1 6.1 6.1 6.1 6.1 6.1 6.1 6.1 3.1 24.4 24.4 24.4
P0
4
[mg/l] nn nn nn nn nn nn nn nn nn nn nn nn nn
L; Cations [meq/l] 0.16 0.12 0.16 0.19 0.19 0.18 0.21 0.23 0.22 0.12 0.6 0.61 0.6
[meq/l] 0.15 0.13 0.16 0.17 0.17 0.16 0.19 0.2 0.19 0.13 0.68 0.68 0.68
153
Table C.4: In situ parameters and ehemieal analysis of ground water sampIes eol-
leeted in the Bedretto-tunnel at sampling loeation Tl (Table C.2; Temp. sampIe
temperature, Cond. electrieal eonduetivity, T DS total dissolved solids, nn not
deteeted).
Parameter SampIes at loeation Tl
Date 14.01.98 05.09.98 13.11.98 10.12.98 24.03.99 27.07.99 04.11.99
Q [l/min] 0.67 0.52 0.41 0.68 0.68 0.75 0.46
Temp. [0G]
19.3 19.3 19.4 19.3 19.2 19.4 19.5
Cond. [JLS/em] 83 82.1 83 83 84.2 82 84.5
pH 9.06 8.93 9.04 8.67 8.97 8.94 8.98
O
2 [mg/l] 3 2.5 2.6
TDS [mg/i] 46.8 47 47.4 47.1 48.2 54.4 50.2
Li [mg/l] 0.01 0.01 0.01 0.01 0.04 0.04 0.05
Na [mg/l] 8.01 8.04 7.88 8.26 8.52 8.47 8.45
K [mg/l] 0.84 0.98 0.92 0.95 0.52 0.63 0.69
Mg [mg/l] nn 0.04 nn nn 0.19 0.19 0.17
Ca [mg/l] 8.5 8.06 8.41 7.95 8.47 8.77 8.26
NH
4
[mg/l] nn nn nn nn nn nn nn
F [mg/l] 3.07 3.29 3.07 3.21 3.74 3.46 3.74
Cl [mg/l] 0.12 0.24 0.15 0.33 0.12 0.12 0.12
Br [mg/l] nn nn nn nn nn nn nn
80
4
[mg/l] 8.29 8.09 8.48 8.27 8.52 8.55 9.45
N0
3
[mg/l] nn 0.23 0.37 nn nn nn 0.92
N0
2
HC0
3
[mg/l] 6.1 6.1 6.1 6.1 6.1 12.2 12.2
P04 [mg/l] nn nn nn nn nn nn nn
~ Cations[meq/l] 0.74 0.75 0.85 0.85 0.88 0.96 0.91
~ Anions [meq/l] 0.8 0.78 0.79 0.78 0.83 0.84 0.81
154
leeted in the Bedretto-tunnel at sampling loeation T2 (Table C.2; Temp. sampIe
temperature, Cond. electrieal eonductivity, T DS total dissolved solids, nn not
deteeted).
Parameter SampIes at loeation T2
Date 14.1.98 5.9.98 13.11.98 10.12.98 24.3.99 27.7.99 4.11.99
Q [l/min] 28.4 25.3 25.7 28.2 25.0 28.2 27.3
Temp. [0G] 19.5 19.8 19.9 19.6 19.8 19.9 19.9
Cond. [JLS/cm] 89 86 88 88 89.9 86 89.5
pH 9.1 9.16 9.29 8.62 9.08 9.01 9.08
O
2 [mg/l] 2.5 2.4 2.1
TD8 [mg/l] 48.6 48 48.5 54.3 50.3 50.5 52.1
Li [mg/l] 0.01 0.01 0.01 0.01 0.04 0.04 0.05
Na [mg/l] 9.17 8.9 8.84 9.17 9.77 9.7 9.71
K [mg/l] 0.82 0.79 0.87 0.77 0.52 0.62 0.71
Mg [mg/l] nn nn nn nn 0.19 nn nn
Ca [mg/l] 7.82 7.74 7.88 7.69 8.64 8.2 8.33
NH
4
F [mg/l] 3.29 3.44 3.37 3.26 3.98 3.75 4
Cl [mg/l] 0.12 0.1 0.09 0.11 0.12 0.19 0.12
80
4
[mg/l] 9.07 8.9 9.02 9.06 9.08 9.45 10.28
N03 [mg/l] nn 0.11 0.27 nn nn 0.57 0.8
N0
2
[mg/l] 0.23 nn nn nn nn nn nn
HC0
3
[mg/l] 6.1 12 6.1 12.2 6.1 12 6.1
I; Cations[meq/l] 0.87 0.77 0.87 0.96 0.9 0.81 0.94
I; Anions [meq/ l] 0.81 0.79 0.8 0.8 0.89 0.85 0.86
155
Table C.6: In situ parameters and chemieal analysis of ground water sampIes co1-
lected in the Bedretto-tunnel at sampling location T3 (Table C.2; Temp. sampIe
temperature, Cond. electrical conductivity, T DS total dissolved solids, nn not
detected).
Parameter SampIes at location T3
Date 5.9.98 13.11.98 10.12.98 14.1.99 24.3.99 27.7.99 4.11.99
Q [l/min] 0.69 0.63 0.64 0.6 0.56 0.58 0.6
Temp. [0C]
20.1 20 19.6 19.9 19.7 20 20.1
Cond. [MS/ern] 69 71 71 71 73.6 74 72.7
pH 9.25 9.34 8.81 9.14 9.02 9.16 9.16
O
2
[mg/I] 3.2 2.9 2.9
TD8 [mg/I] 40.9 42.3 40 40.7 44.8 42.5 44
Li [mg/I] 0.01 0.01 0.01 0.01 0.04 0.04 0.04
Na [mg/I] 6.54 6.83 6.7 6.71 7.2 7.08 7.1
K [mg/I] 1.02 1.06 0.95 0.98 0.85 0.79 0.92
Mg [mg/I] 0.01 nn nn nn 0.19 nn nn
Ca [mg/I] 6.69 7 6.59 6.84 7.31 7.11 7.22
NH4 [mg/I] nn nn nn nn nn nn nn
F [mg/I] 2.28 2.6 2.34 2.31 2.61 2.96 2.93
Cl [mg/I] 0.1 0.07 0.08 0.12 0.18 0.1 0.12
Br [mg/I] nn nn nn nn nn nn nn
80
4
[mg/I] 4.89 5.75 5.02 5.04 5.27 5.13 6.26
N0
3
[mg/I] 1.12 0.73 0.28 0.22 nn 1.1 1.21
N0
2
[mg/I] nn 0.01 0.01 0.27 nn nn nn
HC0
3
[mg/I] 12.2 12.2 12 12.2 6.1 12.2 12.2
P0
4
[mg/I] nn nn nn nn nn nn nn
L; Cations[meq/l] 0.64 0.67 0.63 0.64 0.85 0.68 0.71
L; Anions [meq/ I] 0.65 0.67 0.65 0.66 0.72 0.69 0.7
156
detected).
Date 14.1.98 5.9.98 13.11.98 10.12.98 24.3.99 27.7.99 4.11.99
Q [i/min] 0.87 0.75 0.88 0.94 0.75 0.82 0.94
Temp. [0G] 19.3 19.4 19.3 18.9 19.6 19.2 19.3
Cond. [fLB/em] 94 91 94 94 96 93 96.2
pH 8.85 8.94 9.06 8.91 8.92 8.95 8.91
O
2
[mg/i] 0.9 1 1.5
TD8 [mg/i] 58.3 52.3 63.2 58.9 58.1 61.2 66.3
Li [mg/i] 0.01 0.01 0.01 0.01 0.04 0.04 0.05
Na [mg/i] 8.57 8.37 8.85 8.62 9.37 9.14 9.12
K [mg/i] 0.64 0.69 0.71 0.67 0.45 0.48 0.58
Mg [mg/i] nn nn nn nn 0.2 nn 0.17
Ca [mg/i] 9.39 9.34 9.89 9.35 12.28 9.99 9.97
NH4 [mg/i] nn nn nn nn nn nn
F [mg/i] 3.34 3.53 3.36 3.38 3.96 4.25 3.82
Cl [mg/i] 0.09 0.24 0.08 0.1 0.3 0.13 0.1
Br [mg/i] nn nn nn nn nn nn nn
80
4
[mg/i] 12.27 12.16 12.28 12.32 13.43 12.83 12.53
N0
3
[mg/i] nn nn nn nn nn nn nn
N0
2
HC0
3
[mg/i] 18 12 22 24 6 18 24
P04 [mg/i] nn nn nn nn nn nn nn
2:; Cations[meq/i] 0.93 0.84 1 0.84 1 0.99 1.06
2:; Anions [meq/i] 0.86 0.85 0.9 0.86 1.05 0.91 0.93
157
detected).
Date 5.9.98 13.11.98 10.12.98 14.1.99 24.3.99 27.7.99 4.11.99
Q [i/min] 1.8 2.1 1.9 2.2 2.2 2.6 2.2
Temp. [0G] 19.7 19.8 19.6 19.7 19.7 19.6 19.7
Cond. [JLB/em] 88 88 89 89 91.3 87 91.2
pH 9.07 9.22 9.19 9.06 9 8.96 9.02
O
2
[mg/i] 2.6 2.6 2
TDS [mg/i] 51.2 50.1 50.2 56.9 52.6 53.1 59.6
Li [mg/i] 0.01 0.01 0.01 0.01 0.04 0.04 0.05
Na [mg/i] 8.74 8.18 8.49 8.49 9.02 9.03 9.07
K [mg/i] 1 0.65 0.56 0.61 0.3 0.4 0.51
Mg [mg/i] 0.02 nn nn nn 0.19 0.19 nn
Ca [mg/i] 8.65 8.67 8.56 8.78 9.15 9.28 9.15
NH
4
F [mg/i] 3.51 3.37 3.4 3.4 4.12 4.25 4.1
Cl [mg/i] 0.14 0.11 0.12 0.12 0.13 0.13 0.13
Br [mg/i] nn nn nn nn nn nn nn
S04 [mg/i] 10.87 11.06 11.1 11.15 11.5 11.64 12.34
N0
3
[mg/i] 0.03 0.07 nn nn nn nn nn
N0
2
HC0
3
[mg/i] 12.2 12 12 18.31 12.2 6.1 18.3
P0
4
I; Cations[meq/i] 0.82 0.81 0.81 0.91 0.86 0.97 0.98
I; Anions [meq/ i] 0.84 0.81 0.81 0.82 0.88 0.89 0.87
158
leeted in the Bedretto-tunnel at sampling loeation T6 (Table C.2; Temp. sampIe
temperature, Cond. eleetrieal eonduetivity, T DS total dissolved solids, nn not
deteeted).
Date 5.9.98 13.11.98 10.12.98 14.1.99 24.3.99 27.7.99 4.11.99
Q [l/min] 1.3 1.2 1.3 1.0 1.1 1.1 1.0
Temp. [0C] 19.6 19.7 19.6 19.6 19.6 19.5 19.6
Cond. [ILS/ern] 88 89 89 89 91.3 88 91.6
pH 9.07 9.19 9.25 9.08 8.94 9.09 9.05
O
2 [mg/I] 1.6 1.9 2.2
TD8 [mg/I] 55.5 49.3 49.8 50.2 58.4 57.7 58.8
Li [mg/I] 0.01 0.01 0.01 0.01 0.04 0.04 0.04
Na [mg/I] 9.75 9.66 10.08 10.03 10.82 10.55 10.67
K [mg/I] 0.5 0.54 0.48 0.54 0.28 0.18 0.41
Mg [mg/l] nn nn nn nn nn nn nn
Ca [mg/I] 7.3 7.42 7.49 7.94 8.11 7.85 7.87
NH
4
[mg/I] nn nn nn nn nn nn nn
F [mg/I] 3.09 2.91 2.91 2.95 3.59 3.71 3.59
Cl [mg/I] 0.14 0.1 0.11 0.1 0.13 0.12 0.14
Br [mg/I] nn nn nn nn nn nn nn
80
4
[mg/I] 10.37 10.51 10.65 10.57 11.17 10.97 11.77
N0
3
N0
2 [mg/I] nn nn nn nn nn nn nn
HC0
3
[mg/I] 18.3 6.1 6.1 6.1 18.3 18.3 18.3
P04 [mg/I] nn nn nn nn nn nn nn
I:; Cations[meq/l] 0.88 0.87 0.88 0.88 0.93 0.93 0.94
I:; Anions [meq/l] 0.8 0.81 0.83 0.85 0.89 0.86 0.87
159
Table C.1D: In situ parameters and ehemieal analysis of ground water sampIes
eolleeted in the Bedretto-tunnel at sampling loeation T7 (Table C.2; Temp. sam-
pIe temperature, Cond. eleetrieal eonduetivity, T DS total dissolved solids, nn
not deteeted).
Date 5.9.98 13.11.98 10.12.98 14.1.99 24.3.99 27.7.99 4.11.99
Q [l/min] 7.5 6.6 6.8 7.90 6 6.6 5.4
Temp. [0G]
19 19.1 19 19.1 18.9 18.8
Cond. [JLB/em] 102 104 105 105 105.9 101 105.4
pH 9.13 9.24 9.29 9.15 9.03 9.07 8.98
O
2
[mg/l] 1.9 2.1 1.4
TDS [mg/l] 57.3 63.2 57.2 63.4 66 66.6 66.9
Li [mg/l] 0.01 0.01 0.01 0.01 0.04 0.04 0.04
Na [mg/l] 12.16 12.17 12.6 12.65 13.44 13.4 13.47
K [mg/l] 0.65 0.68 0.56 0.59 0.33 0.37 0.55
Mg [mg/l] nn nn nn nn nn 0.2 nn
Ca [mg/l] 7.68 7.38 7.23 7.29 8.06 8.24 7.88
NH
4
F [mg/l] 3.79 3.61 3.65 3.6 4.4 4.45 4.37
Cl [mg/l] 0.37 0.36 0.35 0.35 0.45 0.45 0.48
S04 [mg/l] 14.44 14.67 14.73 14.65 15.26 15.22 15.78
N0
3
N0
2 [mg/l] nn nn nn nn nn nn nn
HC0
3
[mg/l] 12.2 18.3 6.1 18.3 18.0 12.2 18.3
I; Cations[meq/l] 0.91 1.01 1.01 1 1.06 1.16 1.07
I; Anions [meq/ l] 0.93 0.92 0.92 0.93 1 1.03 1
160
Table C.ll: In situ parameters and chemical analysis of ground water sampIes
collected in the Bedretto-tunnel at sampling location T9 (Table C.2; Temp. sam-
pIe temperature, Cond. electrical conductivity, T DS total dissolved solids, nn
not detected).
Date 5.9.98 13.11.98 10.12.98 14.1.99 24.3.99 27.7.99 4.11.99
Q [l/min] 9.5 11.3 10.4 10.0 10.8 9.4 10.8
Temp. [0G] 15.2 15.8 15.7 15.8 15.4 15.2 15.2
Cond. [flS/cm] 89 92 92 93.1 93.2 88 93.2
pH 9.25 9.37 9.43 9.28 9.13 9.2 9.07
O
2
[mg/l] 1.2 1 1.4
TD8 [mg/l] 53.6 54.1 53.9 54.4 55.6 57.1 57.8
Li [mg/l] 0.01 0.01 0.01 0.01 0.03 0.03 0.04
Na [mg/l] 13.13 13.07 13.53 13.61 14.42 14.52 14.31
K [mg/l] 0.54 0.53 0.46 0.52 0.32 0.33 0.28
Mg [mg/l] nn nn nn nn nn nn 0.17
Ca [mg/l] 4.75 4.72 4.58 5.06 5.03 5.63 5.08
NH
4
F [mg/l] 2.3 2.28 2.37 2.31 2.68 2.76 2.95
Cl [mg/l] 0.14 0.13 0.12 0.13 0.14 0.14 0.16
80
4 [mg/l] 8.42 8.56 8.61 8.6 8.83 8.89 9.68
N0
3
[mg/l] 0.3 0.56 nn nn nn 0.85 1.13
N02 [mg/l] nn nn nn nn nn nn nn
HC0
3
[mg/l] 12 12.2 12.2 12.2 12.2 18.3 18
P0
4
~ Cations[meq/l] 0.9 0.91 0.91 0.9 0.93 0.84 0.87
~ Anions [meq/ l] 0.82 0.82 0.83 0.86 0.89 0.93 0.9
161
Table C.12: In situ parameters and chemical analysis of ground water samples
collected in the Bedretto-tunnel at sampling location Tl! (Table C.2; Temp.
sample temperature, Cond. electrical conductivity, T DS total dissolved solids,
nn not detected).
Parameter Samples at location Tl!
Date 5.9.98 10.12.98 14.1.99 24.3.99 4.11.99
Q [i/min] 3.3 3.3 3.1 2.7 2.7
Temp. [0C] 19.5 19.6 19.6 19.6 19.5
Cond. [tLS/cm] 112 115 115 116.7 116.6
pH 9.3 9.46 9.34 9.2 9.16
O
2
[mg/i] 0.3 0.2 0.4
TD8 [mg/i] 66.9 61 61.3 70.8 64.8
Li [mg/i] 0.02 0.01 0.01 0.04 0.05
Na [mg/i] 15.04 15.5 15.54 16.51 16.52
K [mg/i] 0.78 0.71 0.76 0.46 0.72
Mg [mg/i] nn nn nn nn nn
Ca [mg/i] 6.57 6.48 6.53 7.15 7.06
NH4 [mg/i] nn nn nn nn nn
F [mg/i] 4.99 4.74 4.78 5.69 5.56
Cl [mg/i] 0.61 0.60 0.6 0.74 0.78
Br [mg/i] nn nn nn nn nn
80
4 [mg/i] 14.62 14.86 14.85 15.36 15.99
N0
3
[mg/i] nn nn nn nn nn
N0
2 [mg/i] nn nn nn nn nn
HC0
3
[mg/i] 18.3 12.0 12.2 12.2 6.1
P04 [mg/i] nn nn nn nn nn
I; Cations [meq/i] 1.08 1.08 0.98 1.26 1.15
I; Anions [meq/i] 1 1.02 1.02 1.09 1.2
162
Table C.13: In situ parameters and ehemieal analysis of ground water sampies
eolleeted in the Bedretto-tunnel at sampling loeation SI (Table C.2; Temp. sam-
pIe temperature, Cond. eleetrieal eonduetivity, T DS total dissolved solids, nn
not deteeted).
Parameter SampIes at loeation SI
Date 5.9.98 13.11.98 10.12.98 24.3.99 27.7.99 4.11.99
Q [i/min] 17.8 18.1 18.1 18.5 18.8 19.2
Temp. [0C] 17.6 17.6 17.9 17.7 17.6 17.9
Cond. [JLB/ern] 104 105 106 108.1 102 108.1
pH 9.13 9.2 9.35 9.03 9.17 9.05
O
2 [mg/i] 2.1 1.8 2.2
TD8 [mg/i] 61.9 67.4 62.2 64.2 64.6 71.5
Li [mg/i] 0.02 0.01 0.01 0.04 0.04 0.05
Na [mg/i] 15.07 15 15.49 16.45 16.49 16.49
K [mg/i] 0.8 0.87 0.74 0.55 0.52 0.72
Mg [mg/i] nn nn nn nn nn nn
Ca [mg/i] 5.39 6.62 5.35 5.85 5.79 5.81
NH4 [mg/i] nn nn nn nn nn nn
F [mg/i] 3.24 5.31 3.13 3.42 3.89 3.78
Cl [mg/i] 0.2 0.6 0.15 0.2 0.2 0.2
Br [mg/i] nn nn nn nn nn nn
80
4
[mg/i] 12.87 14.75 13.08 13.47 13.47 14.12
N0
3
[mg/i] 0.10 nn nn nn nn nn
N0
2
[mg/i] nn nn nn nn nn nn
HC0
3
[mg/i] 12.2 12.2 12.2 12.2 12.2 18.3
P04 [mg/i] nn nn nn nn nn nn
~ Cations [meq/i] 1.05 1.2 1.04 1.07 1.09 1.2
~ A n i o n s [meq/i] 0.95 1.01 0.96 1.03 1.03 1.02
163
Table C.14: In situ parameters and chemical analysis of ground water sampIes
collected in the Bedretto-tunnel at sampling location 82 (Table C.2; Temp. sam-
pIe temperature, Cond. electrical conductivity, T DS total dissolved solids, nn
not detected).
Parameter 8amples at location 82
Date 10.12.98 14.01.99 24.03.99 27.07.99 04.11.99
Q [i/min] 28.3 29.1 30 31.1 33.8
Temp. [aC] 18.9 18.9 18.9 18.9 19
Cond. [/LB/ern] 78 80.2 80.4 77 80.4
pH 8.93 8.89 8.9 8.9 8.73
02 [mg/i] 3.3 2.8 2.6
TD8 [mg/i] 58.5 50 54.2 54.7 56.7
Li [mg/i] nn nn 0.03 0.03 0.04
Na [mg/i] 6.93 6.95 7.98 7.45 7.43
K [mg/i] 0.6 0.66 0.62 0.55 0.57
Mg [mg/i] nn nn nn.19 nn nn
Ca [mg/i] 7.57 8.11 7.8 8.3 8.18
NH
4
F [mg/i] 1.71 1.67 1.56 2.09 2.06
Cl [mg/i] 0.16 0.19 0.2 0.21 0.21
Br [mg/i] nn nn nn nn nn
80
4
[mg/i] 11.03 11.08 11.56 11.74 12.46
N0
3
[mg/i] nn nn nn nn 1.44
N0
2
HC0
3
[mg/i] 30.5 21.4 18.3 18.3 18.3
P04 [mg/i] nn nn nn nn nn
1:: Cations [meq/i] 0.82 0.67 0.83 0.86 0.9
1:: Anions [meq/i] 0.69 0.72 0.77 0.76 0.75
Appendix D
Environmental isotopes
164
165
In the following the measured contents in environmental isotopes (tritium,
34Sp80 in aqueous sulphate,
18
0 and 2H) in precipitation and in ground water
sampIes are tabulated. These inc1ude tritium concentrations measured in pre-
cipitation during the 1990's at the meteorological stations Guttannen, Meiringen
and GrimseljHospiz (Table D.2) as weIl as tritium concentrations measured in
two sampling series in the Bedretto-tunnel (Table D.3). Data for tritium content
in precipitation was extracted from the GNIP-database (e.g. Schotterer et al.
(1996)) and additionally provided by the Landeshydrologie, Bern (CH).
34Sp80 content in ground water was measured in one sampling series from
the Bedretto-tunnel (Table DA).
18
0 and 2H content in precipitation was measured in sampIes from the me-
teorological stations Binn, Oberwald, Andermatt, Gtsch and Robiei (Table
D.5). Additionally, 180FH-content in precipitation at the meteorological stations
Guttannen, Meiringen and Grimsel (Table D.6) was extracted from the GNIP-
database and also provided by the Landeshydrologie, Bern (CH). Table D.1 gives
an overview of the location of the meteorological stations.
18
0 FH-content in
ground water was measured in sampIes taken at the sampling locations (Table
C.2) in the Bedretto-tunnel (Table D.7). 180FH-content in glacial ice was de-
termined from ice sampIes taken on the Gerenglacier (Table D.8). For details of
measurement techniques, analytical errors and reference standards refer to Paper
1.
For long measurement records, i.e. tritium and
18
0 FH content in precipita-
tion at the stations Guttannen, Meiringen and Grimsel only exemplary extracts
are tabulated in the following. The complete record can be retrieved from the
attached CD-ROM (Figure D.1).
166
Table D.1: Location of the meteorological stations Guttannen, Meiringen, Grim-
seI/Hospiz, Binn, Oberwald, Andermatt, Gtsch and Robiei. Indicated are sta-
tion number, Swiss map coordinates and altitude of the individual stations.
Station name Station no. X-coord. Y-coord. Altitude
[m] [m] [m a.s.l.]
Guttannen 5030 665290 167600 1055
Meiringen 5070 657570 175550 630
GrimseljHospiz 5010 668460 158160 1980
Binn 7100 657485 135080 1415
Oberwald 7020 669650 154050 1375
Andermatt 4040 688500 165340 1442
Gtsch 4020 690140 167590 2287
Robiei 9296 682540 143984 1903
CD-ROM
\environmental_isotopes\
tritium_prec.dat
-----
d18od2h_prec.dat
Figure D.1: Organisation of data storage on the attached CD-ROM of tritium,
6
18
0 and 6
2
H data in precipitation at the meteorological stations Guttannen,
Meiringen and Grimsel/Hospiz.
167
Table D.2: Monthly mean tritium concentration in precipitation at meteorological
stations Guttannen, Meiringen and GrimseljHospiz (1990's). Tabulated are the
records of two exemplary years. The complete record can be retrieved from the
attached CD-ROM (Figure D.1).
Date Tritium [TU]
GrimseljHospiz Guttannen Meiringen
Jan-95 12.1 9.4 9.3
Feb-95 12.1 11.6 12.4
Mar-95 11.4 13.9 12.4
Apr-95 15.4 17.9 17.9
May-95 23.6 20.7 24.2
Jun-95 25.6 27.4 23.2
Jul-95 19 24.1 26.4
Aug-95 21.2 20 23.8
Sep-95 13.3 10.2 11
Oct-95 19.6 16.4
Nov-95 16.8 17.3 21.7
Dec-95 16.9 15 16.6
Jan-98 7.4 10.1 12.6
Feb-98 7.2 9.2 9.8
Mar-98 9.3 11.1 9.2
Apr-98 10.8 11.7
May-98 19.5 17.5 25.5
Jun-98 14.5 12.9 15.4
Jul-98 24.6 24.2 17.2
Aug-98 9.4 13.3 11.5
Sep-98 9.3 23.1 7.9
Oct-98 9.6 12.6 7.8
Nov-98 10.5 7.6 7.5
Dec-98 13.2 11.4 7.8
168
Table D.3: Tritium concentration in ground water sampIes of two sampling cam-
paigns in the Bedretto-tunnel. For details of sampling locations refer to Table
C.2.
8ampling Ioeation
Tl
T2
T3
T4
T5
T6
T7
T9
Tll
81
82
Tritium [TU]
10.12.98 24.03.99
10.5 9.3
11.7 12.8
13.4 12.5
12.7 12.5
13.6 16.6
8.9 9.0
7.2 6.9
4.4 2.7
3.1 2.6
3.3 2.7
10.5 11.1
Table D.4: 6"
34
8 and 6"
18
0 values in aqueous 8 0 ~ - in ground water sampIes co1-
lected in the Bedretto-tunnel. Date of sampling was 12.7.1999. For details of
sampling locations refer to Table C.2.
8ampling Ioeation
Tl
T2
T3
T4
T5
T6
T7
T9
Tll
81
82
5.79
4.25
6.42
4.51
3.35
5.64
7.57
7.02
9.9
6.8
10.24
-1.57
-4.49
0.81
-5.29
-5.69
-3.18
0.59
2.08
2.75
-2.26
1.88
Table D.5: Monthly mean 6
18
0 and 6
2
H values in preeipitation at the meteorologieal stations Binn, Oberwald, Andermatt,
Gtseh and Robiei. For loeation of the meteorologieal stations refer to Table D.1.
Date Andermatt Oberwald Gtseh Robiei Binn
(\"
18
0
(\"2H
(\"
18
0
(\"2H
(\"
18
0
(\"2H
(\"
18
0
(\"2H
(\"
18
0
(\"2H
[%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0]
Sep 98 -11.6 -82.4 -11.4 -86.6 -13.2 -93.6 -10.1 -70.3 -10.5 -79.1
Oct 98 -13.2 -94.9 -12.1 -89.8 -14.2 -95.8 -12.7 -87.8 -12.7 -92.9
Nov 98 -17.1 -122.6 -15.7 -114.7 -18.1 -123.7 -19.8 -143.8 -17.9 -132.1
Dec 98 -11.2 -80.7 -14.4 -103.0 -20.1 -147.5 -16.1 -114.4 -15.6 -111.4
I
Jan 99 -20.2 -149.4 -15.5 -118.4 -24.2 -176.9 -22.8 -169.8 -21.6 -161.4
Feb 99 -18.3 -133.1 -16.8 -124.6 -26.1 -194.5 -20.8 -152.8 -21.3 -159.0
Mar 99 -13.2 -95.4 -16.9 -127.2 -21.9 -163.5 -17.1 -124.4 -17.0 -125.5
Apr 99 -17.4 -130.1 -16.0 -121.4 -16.4 -122.5 -8.8 -59.9 -10.7 -77.5
May 99 -8.1 -58.1 -12.4 -92.3 -9.6 -65.4 -7.9 -49.4 -7.2 -49.6
Jun 99 -8.1 -54.5 -9.4 -66.7 -8.8 -55.5 -5.8 -33.8 -7.1 -48.0
Ju199 -6.5 -44.5 -8.1 -52.7 -9.7 -65.7 -7.2 -44.8 -7.1 -47.4
Aug 99 -6.9 -45.1 -5.5 -39.8 -8.0 -51.2 -6.6 -42.4 -7.7 -54.4
Sep 99 -10.1 -71.2 -9.5 -67.1 -12.5 -87.3 -9.2 -58.7 -10.5 -72.9
Oct 99 -10.7 -75.5 -12.4 -83.1 -7.1 -40.6 -10.6 -79.2
170
Table D.6: Monthly mean J
18
0 and J
2
H values in precipitation at meteorologi-
cal stations Guttannen, Meiringen and GrimseljHospiz (1970-1999). Tabulated
are the available records of two exemplary years. The complete record can be
retrieved from the attached CD-ROM (Figure D.1).
Date GrimseljHospiz Guttannen Meiringen
8
18
0 8
2
H 8
18
0 8
2
H 8
18
0 8
2
H
[%0] [%0] [%0] [%0] [%0] [%0]
Jan 83 -13.4 -95.4 -13.6 -11.7 -86.9
Feb 83 -18.3 -134.1 -18.1 -17.2 -125.6
Mar 83 -16.0 -117.4 -16.0 -14.3 -104.7
Apr 83 -17.0 -126.4 -14.4 -14.2 -103.6
May 83 -12.5 -88.8 -10.8 -11.5 -82.7
Jun 83 -9.1 -58.2 -8.2 -8.1 -55.1
Ju183 -8.8 -61.2 -6.2 -5.7 -41.2
Aug 83 -8.5 -54.3 -7.3 -6.3 -39.7
Sep 83 -9.9 -64.8 -9.7 -8.7 -57.9
Oct 83 -11.3 -73.3 -10.0 -9.7 -62.6
Nov 83 -13.9 -99.0 -13.0 -12.1 -87.9
Dec 83 -16.9 -129.0 -17.0 -15.5 -113.2
Jan 94 -16.2 -119.1 -13.5 -104.6 -12.3 -90.1
Feb 94 -17.1 -127.9 -19.1 -141.3 -13.6 -103.9
Mar 94 -12.5 -94.1 -11.9 -91.3 -10.6 -76.7
Apr 94 -16.0 -122.0 -14.2 -109.1 -14.4 -115.9
May 94 -12.9 -94.0 -11.2 -82.0 -8.9 -64.8
Jun 94 -9.9 -65.6 -8.3 -58.3 -6.6 -49.9
Ju194 -7.4 -49.3 -5.0 -34.9 -4.3 -29.8
Aug 94 -7.8 -47.9 -5.7 -37.1 -4.9 -28.5
Sep 94 -12.8 -84.5 -10.6 -76.5 -12.4 -88.8
Oct 94 -13.6 -95.3 -12.3 -90.1 -12.2 -92.8
Nov 94 -15.7 -112.0 -15.7 -118.5 -14.9 -111.7
Dec 94 -17.1 -131.2 -16.2 -121.0 -14.7 -109.1
Table D.7: b
18
0 and b
2
H values in ground water sampIes collected in the Bedretto-tunnel. The date of the individual sampling
campaign is indicated. For details of sampling location refer to Table C.2.
Date 04.09.1998 13.11.1998 10.12.1998 14.01.1999 24.03.1999 27.07.1999 05.09.1999
15
18
0 J
2
H 15
18
0 J
2
H 15
18
0 J
2
H 15
18
0 J
2
H 15
18
0 J
2
H 15
18
0 J
2
H 15
18
0 J
2
H
[%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0] [%0]
Tl -14.89 -105.60 -14.98 -106.00 -15.02 -106.80 -15.00 -106.75 -15.08 -106.60 -15.06 -106.80 -15.01 -106.40
T2 -14.96 -105.70 -14.92 -105.45 -14.94 -106.05 -15.01 -106.45 -15.04 -106.60 -15.03 -106.40 -14.95 -106.30
T3 -14.84 -104.90 -14.84 -105.30 -14.91 -105.70 -14.88 -105.50 -14.93 -105.90 -14.93 -106.40 -14.84 -105.30
I ~
T4 -14.77 -104.60 -14.80 -104.90 -14.80 -105.20 -14.83 -105.20 -14.79 -104.60 -14.83 -105.70 -14.72 -104.80
T5 -14.73 -104.30 -14.69 -103.90 -14.69 -104.30 -14.71 -104.30 -14.77 -105.20 -14.70 -105.00 -14.74 -105.30
T6 -14.85 -104.70 -14.85 -104.95 -14.86 -105.40 -14.87 -105.20 -14.90 -105.40 -14.95 -106.50 -14.87 -105.70
T7 -14.71 -104.00 -14.66 -103.55 -14.69 -104.45 -14.70 -104.35 -14.76 -103.80 -14.73 -103.90 -14.65 -104.20
T9 -14.89 -104.90 -14.87 -104.20 -14.92 -105.05 -14.96 -105.15 -15.07 -105.90 -14.94 -104.50 -14.86 -104.50
Tll -14.81 -104.60 -14.81 -105.15 -14.83 -105.25 -14.87 -105.60 -14.80 -104.90
SI -14.79 -104.30 -14.82 -104.45 -14.78 -104.40 -14.81 -104.30 -14.85 -104.50 -14.83 -104.30 -14.84 -104.20
S2 -15.00 -106.25 -14.98 -105.70 -15.00 -106.10 -15.04 -105.90 -14.93 -105.90
172
Table D.8:
18
0 and
2
H values in ice sampIes collected on the Gerenglacier.
Indicated are the Swiss map coordinates of the sampling Iocations.
Sampling Date
18
0
2
H X-coord. Y-coord.
Iocation [%0] [%0] [m] [m]
gl 29.8.99 -13.81 -100.9 678150 153260
g2 29.8.99 -13.78 -101.1 678560 153360
g3 29.8.99 -13.53 -94.6 678230 153050
g4 22.9.99 -12.73 -88.3 678660 153330
g5 22.9.99 -13.38 -94.2 678400 153050
Appendix E
Discharge measurements
173
174
In the following, discharge measurements in the Bedretto- and Furka-base-
tunnel (Table E.l and E.2), carried out in the basal drain of the galleries on
a monthly/bi-monthly basis to characterize ground water inflow along specific
sections of the galleries, are tabulated. For details on the characterized tunnel
sections and on the measurement technique refer to Paper 1 and 2.
The continuous records of the discharge measurements at the individual sta-
tions 81 and 82 in the Bedretto-tunnel (Figure 2.2) are stored on the attached
CD-ROM (Figure E.l). For locations and details on these monitoring sites refer
to Paper 1 and Table C.2.
Table E.l: Integral flow rates to the Bedretto-tunnel (total accessible section)
and to the fault zone section within the Bedretto-tunnel together with integral
flow rates to the gallery in the upstream of individual measurement locations in
the basal drain along the Bedretto-tunnel (referenced to tunnel metries; refer to
Figure 3.4 and 3.2). Associated measurement error is O.5 l/s.
Date Bedretto-tunnel Fault zone >TM4128 >TM4325 >TM4510
Q [li s] Q [li s] Q [li s] Q [lls] Q [lls]
May 99 14.4
Jun 99 14.6
Jul99 14.5
Sep 99 14.5
Oct 99 14.7
Nov 99 15.1 7.4 14.3 6.9 4.7
Dec 99 14.8 7.6 14.3 6.7 3.8
Jan 00 14.8
Feb 00 14.7 7.4 14.2 6.8 4.3
Apr 00 15.0 7.6 14.3 6.7
May 00 15.0 7.2 14.3 7.1 4.0
JulOO 15.3 7.5 14.6 7.1 3.7
Aug 00 15.3 7.5 14.8 7.3 3.9
Sep 00 15.4 7.4 14.7 7.3 4.1
Nov 00 15.4 7.4 14.7 7.3 4.0
175
Table E.2: Integral flow rates to the Rotondo granite section of the Furka-
basetunnel (RG) and total discharge at the tunnel portals in Oberwald and Realp.
For general direction of tunnel drainage refer to Figure 3.4. Associated measure-
ment errors are 5.6 l/s and 5.9 1/s for the granite section and the Portal
measurements respectively.
Date Furka-basetunnel (RG) Portal Realp PortalOberwald
Q [l/8] Q [l/8] Q [l/8]
Sep 98 27.6 96.4
Oct 98 25.2 75.8 94.7
Nov 98 25.1 75.6 94.9
Dec 98 27.3
Jan 99 26.1 75.8 94.8
Jun 99 24.5
Ju199 76.0 94.4
Aug 99 24.6 93.6
Sep 99 25.0 93.3
Nov 99 25.5 93.1
Dec 99 94.3
Jan 00 27.1
CD-ROM
\d ischarge\
Figure E.1: Organisation of data storage on the attached CD-ROM for the records
of continuous discharge measurements in the Bedretto-tunnel at locations 81 and
82 (Figure 2.2).
Appendix F
Dimensionless solution for
constant head test in ideal
aquifer with a well of finite
radius and constant head
boundary
176
177
The dimensionless solution in the Laplace domain for a constant head test
in a finite radius (rw) weIl (i.e. tunnel) with an infinite linear constant head
boundary at a dimensionless distance h
d
has been derived by convolution as:
with
qd: dimensionless drawdown
p: Laplace parameter
K
o
, K
1
: modified Bessel function of 2nd kind of order 0 and 1 respectively
(F.1)
To produce figures 3.22 and 3.23 the solution was inverted using the Stehfest
algorithm (Stehfest 1970).
Curriculum Vitae
Personal Details
Name:
Firstname:
Date of birth:
Place of birth:
Nationality:
Education
1977-1981
1981-1987
1987-1988
1988-1990
12.06.1990
University studies
10/1991-09/1993
27.08.1993
09/1993-07/1994
10/1994-12/1995
Ofterdinger
Ulrich Stefan
18. November 1970
Kassel (FRG)
German
Elementary school Am Heideweg, Kassel (FRG)
Gymnasium Engelsburg, Kassel (FRG)
Student exchange to New Zealand,
Rangitoto College, Auckland (NZ)
Gymnasium Engelsburg, Kassel (FRG)
Final degree: Abitur
Studies in Geology at the Philipps-Universitt,
Marburg/L. (FRG)
Pre-diploma in Geology
Studies in Geology at the University of Liverpool,
Liverpool (UK)
Studies in Geology at the Christian-Albrechts-
Universitt Kiel (FRG)
178
01/1996-01/1997
02/1997-07/1997
10.07.1997
08/1997-02/2001
179
Accomplishment of Diploma thesis (research and mapping
project) at the Rijksuniversiteit Gent, Ghent (B)
Mapping: Cenozoic sediments in the Flemish Ardennes,
Zwalm-region (B)
Research: Comparison of field and laboratory determi-
nation of the hydraulic conductivity of natu-
ral and artificial liner materials
Final Diploma examinations at the Christian-Albrechts-
Universitt, Kiel (FRG)
Final degree: Diploma in Geology
PhD-position at the Geology Institute/Engineering
Geology, ETH-Zrich (CH)

Ofterdinger 2004 Ground Water Flow Systems Rotondo Granit ETH 14108

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ofterdinger 2004 Ground Water Flow Systems Rotondo Granit ETH 14108

Uploaded by

Copyright:

Available Formats

I

3.4 Conceptual model

You might also like