Physics and Chemistry of the Earth 64 (2013) 127–139

Contents lists available at SciVerse ScienceDirect

Physics and Chemistry of the Earth

journal homepage: www.elsevier.com/locate/pce

Halite clogging in a deep geothermal well – Geochemical and isotopic

characterisation of salt origin
Annalena Hesshaus ⇑, Georg Houben, Robert Kringel
Federal Institute for Geosciences and Natural Resources, Stilleweg 2, 30655 Hanover, Germany

a r t i c l e i n f o a b s t r a c t

Article history: The sandstone formation of the Middle Buntsandstein (Lower Triassic) in the geothermal well Groß
Available online 5 July 2013 Buchholz Gt1, Hanover, Northern Germany, was hydraulically stimulated to generate a heat exchanger
surface, using 20000 m3 of fresh water. After six months of enclosure the recovered water was oversat-
Keywords: urated with respect to halite at surface conditions. Due to cooling induced precipitation a salt plug
Geothermal well formed between 655 and 1350 m depth in the tubing. While the Na/Br and the Cl/Br ratio of the recovered
Halite water reflect the signature of a relic evaporative solution the recovered water contains tritium, indicating
Clogging
a significant proportion of fresh water. Leaching experiments of the reservoir rocks point towards pres-
North German Basin
Hot water rock interaction
ence of traces of soluble salt minerals in the formation. Therefore we assume that the salinity cannot be
attributed solely to halite dissolution nor to the production of a pure formation brine. The recovered
water is a result of a combination of both salt dissolution by injected fresh water and of mixing with a
formation brine which has undergone water–rock interaction. The calculated fresh water proportion in
the recovered water is around 40%. The presence of salt mineral traces in pores of a target formation is
a potential threat for the operation of geothermal wells, as cooling-induced salt scaling jeopardizes their
performance.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction
Geothermal energy is a very promising source of renewable
energy which is being investigated in many parts of the world.
Deep sedimentary basins such as the Northern German Basin
(NGB) provide potential target horizons.
The NGB consists of mostly horizontally bedded sediments of
Mesozoic and Cenozoic age. Sediments of suitable temperature
for geothermal exploitation often have low to negligible porosity
and consequently low permeability. To develop and test possibili-
ties to generate geothermal energy from these sediments of low
permeability, the GeneSys project (generated geothermal energy
systems) was initiated in 2003 (Kehrer et al., 2005).
The project intends to provide thermal energy (2 MWth) for a
building complex in Hanover (Fig. 1). For this purpose, the geother-
mal well Groß Buchholz Gt1, approximately 4 km deep, was drilled
in 2009. The well was prospected to reach the sandstones of the
Middle Buntsandstein (Lower Triassic) (Fig. 2). The innovative as-
pect of the project was the ‘‘single-well’’ approach for producing
from the Triassic and injecting the thermal water in the more shal-
low Cretaceous sandstone through the annulus for storage (Kehrer
et al., 2005).
⇑ Corresponding author. Tel.: +49 5116432613.
E-mail address: Annalena.Hesshaus@bgr.de (A. Hesshaus).
Because of the very tight (<1 mDarcy) sediments of the Middle
Buntsandstein it was necessary to generate an artificial fracture as
an underground heat exchanger. The technique was tested in a re-
search project at the deep well Horstberg Z1 (Fig. 1) which is lo-
cated approximately 80 km north of Hanover. The stratigraphic
profile of Horstberg Z1 is similar to the one of Groß Buchholz
Gt1 and its production horizon is the same (Tischner et al.,
2010). Results from analysis of the formation brine were available
from several production tests.
With this experience and the results of a pretest in Gt1 an
artificial fracture at Groß Buchholz Gt1 was generated in 2011,
injecting 20,000 m3 of fresh water into the well. After six months
of enclosure the water was recovered in order to deduce hydraulic
and hydrochemical data of the fracture and the reservoir
conditions. During the tests, the injected and produced water
was sampled in intervals. The comparison of the results allowed
characterising the geochemical properties of the reservoir.
Immediately after exchanging the water volume of the tubing,
the salinity of the produced water increased abruptly. The water
became oversaturated with respect to halite at production condi-
tions, so that several parts of the above-surface equipment clogged
and failed immediately. Intense cooling occurred in the production
tubing, so that a salt plug developed in 655 m depth, which clogged
the well. The production test had to be stopped. One year after the
recovery test the salt plug was removed by injecting fresh water

1474-7065/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.pce.2013.06.002
128 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139

with a capillary coil unit (Multiline). With interruptions the salt

Fig. 1. Map of Germany with regions with geothermal potential (shaded, after
Schulz et al., 2007). Geothermal wells Groß Buchholz Gt1 and Horstberg Z1 are
marked with open circles.
plug went down to 1350 m.
Prior to the injection of fresh water it was not possible to sam-
ple the originally present formation brine due to the very low per-
meability. The porosity of the formation rocks ranges between 1
and 2 vol.% (pers. comm. J. Orilski, Leibniz-Institute for Applied
Geophysics (LIAG), 2012). In spite of this, the rocks are most likely
watersaturated with no free gas phase.
Fig. 2 shows the geology of the well Groß Buchholz Gt1, as well
as the well completion and an estimate of the dimensions of the
artificial fracture. The real fracture dimensions are still unknown
due to insufficient hydraulic data. The estimate shown is based
on rock mechanical model pre-calculations (Tischner et al.,
2013). According to these, the fracture has a plane of around
0.6 km2 with a maximum length of 2000 m, maximum height of
400 m and maximum width of 3 cm at the injection point. It is
noteworthy that strata containing rock salt (footwall: Zechstein
formation, headwall: Röt formation, Upper Buntsandstein) are
present in the proximity of the fracture.
Because no capacities to store the produced water above ground
were available, it was reinjected into a more shallow storage hori-
zon. Based on previous investigations, discussed by Schäfer et al.
(2012), a sandstone formation in the lower Cretaceous (Wealden
sandstone) at 1200 m depth was selected.
In this paper we compare the compositional changes between
injected and recovered water, both for the pretest and the fractest.
In addition, we compare the results with a test at Horstberg Z1 in
the same formation. The aim is to infer the geochemical in situ

Fig. 2. Stratigraphic profile and sketch of well completion of geothermal well Groß Buchholz Gt1. The target horizon lies in approximately 3750–3400 m and the storage
formation in approximately 1200 m. Right: zoom into the target horizon, the likely fracture dimensions are displayed as side face. The fracture plane proceeds perpendicular
to the drawing area. The depth of the Zechstein top was interpolated from seismic profiles and after data from Baldschuhn et al. (2001).
 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139 129

Fig. 3. Operational parameters of the main fractest.

properties in the target horizon and the origin of the salt clogging.
To solve these questions, water, gas, and solids were analysed for
their hydrogeochemical, stable isotope and radionuclide
composition.
2. Methods
2.1. Injection and recovery tests
The tests considered here comprise a small pretest and the main
fractest. The objective of the first was to determine the quality of
the tubing perforation. For the pretest, performed in April 2011,
some 100 m3 of fresh water were injected into the well and recov-
ered after five days of enclosure. During this test no large fracture
was generated.
Fig. 3 shows the operational parameters of the main fractest,
like injection rates and pressure, both displayed as a function of
time. For the generation of the fracture, 20,000 m3 of filtrated fresh
water from the Mittelland Channel (MLC), an artificial shipping
lane, were injected into the formation in five steps over a time per-
iod of five days. The overnight enclosure periods show only a small
pressure decrease, which can be related to minor losses of water.
Therefore the presence of a pervious fracture could be assumed
(Tischner et al., 2013). After almost six months of enclosure recov-
ery of the injected water was commenced in November 2011
(Fig. 4). During the first production phase almost 140 m3 of water
was produced, and during the second phase almost 440 m3, in total
580 m3 were recovered.
Both recovery tests were operated under high pressure (max.
400 bar). The water flow was led through high pressure pipes in
a flow-through loop from the production site to the injection site,
where it was injected directly into the storage formation
(Wealden) via the casing annulus (Fig. 2). A 50 lm filtering system
and several monitoring instruments were installed in the flow-
through loop, including sensors for temperature, pressure, density,
and flow rate. A sampling loop and a separate sampling tab were
installed as well. In order to prevent fracturing in the Wealden
formation, the injection pressure was restricted to a maximum
value of 188 bar at formation depth, resulting in a flow rate of
2–4 l s1, due to requirements of the local mining authority. The
low flow rate promoted cooling of the ascending water during
the passage through the tubing. Therefore the maximum tempera-
ture of the recovered water at the well head reached only values of
around 50 °C while the measured formation temperature was
almost 170 °C.
The recovered water showed oversaturation with respect to ha-
lite at well head conditions, immediately after exchange of the well

Fig. 4. Operational parameters of the recovery test after the fractest (sm = Middle Buntsandstein).
130 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139
volume. Water salinity and the subsequent precipitation caused
increasing injection pressure in the Wealden formation, so that
the limit for reinjection mentioned above was reached and the
recovery had to be interrupted. To redissolve precipitates, fresh
water was injected into the annulus. The intense cooling by fresh
water further promoted scaling in the tubing, producing a salt plug
at 655 m depth. The recovery test thus had to be stopped. There-
fore only a small fraction of the injected water could be recovered.
As a consequence, it was not possible to gain sufficient flow and
pressure information to derive fracture dimensions.
2.2. Sampling and sample treatment
The sampling equipment was not designed to handle oversatu-
rated water. Consequently, filters and pressure reducer for the
sampling loop were blocked shortly after production of the highly
saline water. The sampling loop was therefore not useable and
sampling had to be conducted from a stainless steel sampling tap
installed into the flow-through loop.
The water was sampled for analysis of major and trace elements
and for its isotopic composition. To analyse the total inorganic and
organic carbon, a 20 ml glass vial was filled bubble-free. For cations
and trace elements a 100 ml HDPE bottle containing 1 ml HNO3
(65%) was filled with filtrated (0.45 lm, cellulose acetate) water
immediately. For anions, alkalinity, ammonium, density and the
isotopic composition the water was filled bubble free into a
500 ml HDPE bottle.
For the determination of the physico-chemical parameters, like
specific electrical conductivity, pH, temperature and oxidation
reduction potential the water was run through an overflowing
bucket, in order to minimise air contact. The preferable use of a
flow-through cell was impossible due to the particle-rich and
gas-loaded water.
A retain sample of 0.5 m3 was filled into a closed container. The
samples for analysis of isotopic composition of dissolved sulphates,
of tritium activity and activity of other natural radionuclides were
taken from this batch.
The injected water for the fractest was taken from the MLC near
to the well location. All samples which were not taken during
injection were collected directly from the channel afterwards, as
it was done for the analysis of the tritium activity and for the iso-
topic composition of dissolved sulphates.
During both recovery tests the exsolved gas was sampled. It was
intended to separate water and gas with a gas separator (Seeger,
2011) to determine the water to gas ratio as described in Schröder
and Hesshaus (2009). The unit was separated from the high pressure
flow-through loop by a pressure reducer. Immediately upon pro-
duction of the halite oversaturated water, the reducer was blocked
and hence failed. Therefore it was necessary to sample the exsolving
gas with a bottle held overhead in an over-flowing bucket. Conse-
quently it was not possible to measure the water to gas ratio.
During the drilling of Groß Buchholz Gt1 in 2009 the reservoir
rocks of the Middle Buntsandstein formations (Detfurth and Volp-
riehausen series) were cored (Schäfer et al., 2012). In total almost
40 m of core were available, amounting to approximately 10% of
the probable fracture height of 400 m. Samples were taken over
the entire length of the cored intervals, to determine the petrogra-
phy of the rocks (Röhling and Heinig, 2012) and to analyse the
mineralogical and geochemical composition. These rock samples
were, at least partially, in contact with saline drilling fluid and
water during sample preparation. To determine the content of
highly soluble minerals, like sodium chloride, additional rock sam-
ples were taken without water contact. For this purpose, cuttings
of drill cores were collected by drilling with a 10 mm hard metal
drill. 15 samples of 250 mg of the cuttings were eluted with
50 ml distilled water, shaken overhead for approximately 72 h,
filtrated with a 0.45 lm cellulose acetate-filter and the superna-
tant was analysed.
The salt plug was sampled with a wire line tool (bailer) for sub-
sequent analysis of its chemical and mineralogical composition as
well as its activity of natural radionuclides.
2.3. Analytical methods
Water samples had to be diluted 1:200 with distilled water. For
analysis of fluoride, sulphate and nitrite an ion chromatograph ICS
3000 (Dionex) was used. The alkalinity was determined by titra-
tion. Ammonium was determined using a photometer Unicam
UV300 (Thermo Electron Corporation). Cations, chloride, bromide
and sulphate were analysed via inductively coupled plasma optic
emission spectrometry using a Ciros by Spectro. The total inorganic
and organic carbon was measured by infra red detection with a
High TOC II (Elementar). The trace metals were analysed using
inductively coupled plasma mass spectrometry with a 7500ce by
Agilent. The density was measured with an oscillating U-tube
DMA 38 (Anton Paar).
The isotopic composition of water was analysed with a cavity
ringdown spectrometer type L2120-i (Picarro). The sulphate iso-
topes of both water and rock were analysed at the University of
Münster, Germany, with element analyser isotope mass spectrom-
eter Delta Plus (ThermoFinnigan) for d34S and with a combination
of high temperature pyrolysis (ThermoFinnigan thermal combus-
tion element analyser) and Delta Plus XL mass spectrometer (Ther-
moFinnigan) for d18O. The d13C-ratio of carbonate cements of the
reservoir rocks was determined with isotope ratio mass spectrom-
etry, using a Deltaplus XP (Finnigan). The tritium activity was mea-
sured at VKTA Rossendorf e.V. (Nuclear Engineering and Analytics
Inc.) in Dresden, and by Hydroisotop GmbH, Schweitenkirchen,
Germany, using liquid scintillation spectrometry (Quantulus spec-
trometer). Tritium activity was also determined at the University
of Bremen with a noble gas mass spectrometer type MAP215-50
(MAP). The composition and activity of naturally occurring radio-
active material were determined by the Institute of Radioecology
and Radiation Protection (IRS) of the Leibniz University of Hanover.
For analysis three detectors were used (GR2817 and GX3018 by
Canberra and GEM-20190 by Ortec).
The chemical composition of the gas phase was analysed with
micro gas chromatography using a Varian CP4900.
The salt plug material was dissolved in distilled water and ana-
lysed equivalent to the procedure for the water samples described
above. In addition, solid salt was investigated with scanning elec-
tron microscopy (SEM) with a FEI Quanta 600F. Insoluble residues
of the salt plug, as well as rock samples, were analysed with X-ray
diffractometry with an X0 Pert Pro H-2 H (PANalytical). The chem-
ical composition of the rock samples was investigated with X-ray
fluorescence analysis with a PANalytical Axios and a PW2400.
2.4. Thermodynamic calculations
The thermodynamic calculations were done with the program
PhreeqC (version 2.18.00; Parkhurst and Appelo, 1999) using the
phreeq (Parkhurst and Appelo, 1999) and Pitzer database (Pitzer,
1981).
3. Results
3.1. Reservoir rocks
The target horizon for the geothermal well Groß Buchholz Gt1 is
located between 3420 and 3843 m true vertical depth, from which
four cores were taken (Schäfer et al., 2012). The formation is an
 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139 131

3.2. Injected and recovered water

The injected water was nearly saturated with respect to oxygen

and contained considerable concentrations of nitrate and nitrite
(Table 3).
The water recovered during the pretest is a very hard Na–Ca–Cl
1
brine (Table 3). With a total dissolved solids (TDS) of 302 g kgw
and a density of 1.21 g cm3 (25 °C), it is less mineralised than
the recovered water of the fractest. After degassing the pH was just
under 7. Relatively high concentrations of ferrous iron and manga-
nese (II) indicated reducing redox conditions.
The recovered water of the fractest is also a very hard Na–Ca–Cl
1
brine with a TDS of at least 452 g kgw (Table 3). This value was cal-
culated from bulk chemistry, corrected for cooling induced precip-
itation during sampling and sample preparation. All samples taken
during the recovery of the fractest are considered equivalent taking
into account the accuracy of measurement. The calculated density
amounts 1.23 g cm3 at 25 °C, compared to the measured density
Fig. 5. Content of chloride, sodium and sodium plus potassium dissolved from rock
samples compared to the content that could be generated by dissolution of dried
of 1.22 g cm3. This difference can be attributed to the precipita-
formation brine or drilling fluid. tion. The pH of the water was determined after partial degassing
at production temperatures to lie between 5.5 and 5.7 at surface
conditions. With high concentrations of ferrous iron, manganese
alternating sequence of silty sandstones with varying amounts of (II) and ammonium (Table 3) the redox system of the water is
clay minerals and clayey siltstones and claystones. The sedimenta- considered reducing. In all samples the analysed sulphate concen-
tion environment was terrestrial with varying (lacustrine, wet trations were far lower than concentrations for total dissolved
sandflat, mudflat and dry sandsheet) conditions (Röhling and Hei- sulphur. A significant proportion of dissolved sulphur must there-
nig, 2012). All samples consisted of quartz, feldspar, and mica in fore be present as meta-stable sulphur species, in the form of inter-
the order of decreasing percentage and in the case of the clayey mediate species between sulphide and sulphate, e.g., thiosulphate
siltstones of clay minerals like illite and muscovite. The cements or sulphite. Therefore a sulphide oxidising redox system can be
consisted of silica, calcium and magnesium carbonates, calcium assumed (Appelo and Postma, 2005).
sulphates and iron oxides. The elution experiments showed that Due to cooling-induced precipitation some of the analysed con-
sodium chloride can be dissolved from the reservoir rocks centrations were minimum concentrations. Especially for sodium
(Fig. 5). For 24 of 30 samples the amount of dissolved sodium and chloride this is to be considered in detail, because of halite pre-
and chloride was higher than it could be generated by simple cipitation during the recovery test, i.e., salt plug.
dissolution of dried formation brine with a porosity of 2 wt% The water of the fractest was investigated for its isotopic com-
1 1
(0.04 mol kgrock ; 2.3 g kgrock ). Therefore halite minerals with position and its radionuclide activity (Table 2). Fig. 6 shows the
1
amounts up to 0.7 wt% (6.6 g kgrock ) and a mean value of 0.2 wt% mean values of isotopic ratio of deuterium and oxygen-18 of the
1
(2.1 g kgrock ) must be present in the pore spaces. The supernatant injected and the recovered water. As expected, the injected water
contained potassium concentrations exceeding those of the recov- plots close to the global meteoric water line (GMWL). The recov-
ered water of the fractest. This is most probably an artefact, caused ered water deviates strongly from the GMWL. It is isotopically hea-
by potassium release from illite and muscovite during the experi- vier than the injected water, with a 20‰ increase in d2H and a
ment. The sum of sodium and potassium matches the concentra- 6.5‰ increase in d18O.
tion of chloride very well. In Table 1 the reactive phases of the The tritium activity of the recovered water lies in the range of
reservoir rocks are displayed. The high iron content in some 3.2–3.3 TU, compared to 8.5 TU in the injected water (Table 2).
samples can be attributed to iron oxide cementation. Heinig et al. The percentage of fresh water in the recovered water is of high
(2011) and Röhling and Heinig (2012) provide more detailed importance. For this purpose the tritium activity (Table 2) was
descriptions of petrology and geochemistry of the reservoir rocks. used. A conservative calculation of a two-member mixing (Eq.
Only the bulk composition of the isotopes of sulphate and the (1)) was performed.
carbonate cements were measured (Table 2).
xA þ yB ¼ M ð1Þ
with A is tritium activity of end-member A (formation brine), B is tri-
Table 1
Content of reactive phases of the reservoir rocks (Buntsandstein) of Groß Buchholz tium activity of end-member B (injected fresh water), M is tritium
Gt1 (n = 25). activity of mixing product, x is percentage of A, y is percentage of B.
Following this approach the fresh water percentage lies in the
Min. content Median content Max. content
(g kg1) (g kg1) (g kg1) range of 40% (Fig. 6).
C (carbonate) <0.1 2 36
C (org.) <0.1 <0.1 <0.1 3.3. Calculation of stable isotope ratios in the formation brine
S (sulphate) <0.1 3 20
Al 9 22 43 Under reservoir conditions fractionated d18O equilibrium be-
Fe 1.1 4.3 19 tween the anhydrite and water is expected. The following Eq. (2)
Mn 0.01 0.2 1.3
(Clark and Fritz, 1997) allows an estimate of the isotope ratio of
Mg 1.6 6.4 19
Ca 2.9 14.2 122 oxygen of the hypothetical formation brine.
Ba 0.25 0.69 2.9
Pb 0.004 0.01 0.02 1000 þ d18 OSO4
Sr 0.11 0.14 0.3
d18 OH2 O ¼  1000 ð2Þ
106
expða  T2
þ bÞ
132 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139

with a is empiric parameter (dimensionless), b is empiric parameter

2–12
(±‰)
2.5 (dimensionless), T is temperature (K).
SD

Published values for a and b vary, Chiba et al. (1981) proposed

a = 3.21 and b = 4.72, Lloyd (1968) 3.25 and 5.1, and Mizutani
Tritium

3.2–3.3
(n = 3)

(n = 2)
and Rafter (1969) cited in Clark and Fritz (1997) 2.88 and 3.6.
(TU)
8.5

Despite these differences at T = 170 °C (=443.15 K) and a mean

d18 OSO4 of 15.6‰, the resulting d18 OH2 O signature is expected to fall
(±‰)

between 3.9‰ and 4.4‰ VSMOW.

0.4

0.4

0.4
SD

A similar calculation can be performed for carbonate cements

with Eq. (3) (O’Neil et al., 1969):
(VSMOW) (‰)
Stable isotope composition and tritium activity of injected and recovered water of the fractest and of carbonate and sulphate cementation minerals of the reservoir rocks. SD = standard deviation.

1000 þ d18 OCaCO3

d18O carb.

d18 OH2 O ¼ 6
 1000 ð3Þ
expð2:78  10  3:39Þ
15.1

17.8

19.1
T2

For T = 443.15 K and d18 OCaCO3 VPDB = 13.2‰ (mean value; d18-
O CaCO3 VSMOW = 17.3‰) d18 OH2 O would be 23.8‰ VPDB (Vienna
(PDB) (‰)
d18O carb.

Pee Dee Belemnite) equivalent to 6.4‰ VSMOW (Clark and Fritz,

15.18

12.62

11.93

1997).
The hydrogen isotope composition of hypothetical formation
brine was calculated using Eq. (4) after Gilg and Sheppard
(±‰)

(1996). The isotope composition of clay minerals in sedimentary

0.2

0.2

0.2
SD

rocks as given by Hoefs (2009) covers a wide range from 25‰

to 150‰ VSMOW. With this range of input data, the resulting val-
(PDB) (‰)
d13C carb.

ues of d2 HH2 O range from 6.4‰ to 133.8‰.

1.92

0.74

0.13

1000 þ d2 Hclay minerals

d2 HH2 O ¼ 6
 1000 ð4Þ
expð2:2  10
T2
 7:7Þ
(±‰)
0.2

0.2
SD

Fig. 6 shows that the range of the calculated ratios of the stable
isotopes of the formation brine (lightly shaded box) intersects with
(VSMOW) (‰)

the dashed mixing line.

d18O H2O

The lightly shaded box in Fig. 6 shows the range in which the
0.15

isotope composition of the formation water would fall based on

6.7

calculations (Section 3.7) after Chiba et al. (1981) and Gilg and
Sheppard (1996). The darker shaded box shows the measured
(±‰)

d18O ratio of sulphate and carbonate cements of the reservoir

SD

rocks.
(VSMOW) (‰)

3.4. Exsolved gas

d2H H2O

29.9
50.1

The composition of the exsolved gas of the pretest water

showed little variation. Out of four samples, mean values could
(±‰)

be calculated with small standard deviation. After subtraction of

0.2

0.5

0.2

0.2

0.1
SD

air contamination, the most important fraction of the gas is nitro-

gen (almost 96 vol.% ± 0.2). Minor components are hydrogen
(VSMOW) (‰)

(3.9 vol.% ± 0.2), methane (0.016 vol.% ± 0.001), and carbon dioxide
(0.003 vol.% ± 0.001).
d18O SO4

The compositions of the three exsolved gas samples of the frac-

11.7

15.2

15.4

16.2
5.4

test on the other hand vary considerably. All samples consist

mostly of nitrogen (82.8–93.7 vol.%) and in minor parts of carbon
(±‰)

0.04

dioxide as well as traces of hydrogen (0.5–1 vol.%) and methane

0.2

0.2

0.1

0.1
SD

(0.01 vol.%). A shift in composition, after subtraction of air contam-

ination, can be observed with the increase of carbon dioxide from
(VCDT) (‰)

5.2 vol.% CO2 in the first to 16.7 vol.% CO2 in the last sample. Be-
d34S SO4

cause of the problems mentioned above the water to gas ratio

6.8

13.7

13.5

12.6
9.8

could not be determined.

Injected water MLC fractest

3.5. In situ pH
Reservoir rocks carbonate

Reservoir rocks carbonate

Reservoir rocks carbonate

Recovered water fractest

Reservoir rocks sulphate

Reservoir rocks sulphate

Reservoir rocks sulphate

cements sample 1

cements sample 2

cements sample 3

cements sample 1

cements sample 2

cements sample 3

The pH of the recovered water (5.5–5.7) was measured after

(measured) (4)

degassing, prior to the oxidation of iron at a temperature of

25 °C. To infer the in situ pH, equilibrium with carbon dioxide
(e.g., partial pressure of CO2, 16.7 vol.%) was assumed and then
Sample

approximated with thermodynamic modelling (PhreeqC), keeping

(3)
Table 2

in mind the insufficiency of both thermodynamic database and

model. At a temperature of 100 °C an in situ pH of around 4 would
 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139 133

Table 3
1
Chemical composition of injected and recovered water of Groß Buchholz Gt1 and Horstberg Z1. Concentrations in mg kgw .

Sample Density Sampling pH (–) EC EH TDS TDS Ba BO2 Br Ca Cl Fe HCO3 K Li

(25 °C) temp. (°C) (lS cm1) (mV) (g L1) (g/kg1 )
w
(g cm3)
Injected water 1 13.5 7.5 512 n.a. 0.4 0.4 0.06 0.1 0.06 71 40 0.06 128 2.7 0.003
pretest (2)
Recovered water 1.21 33 6.96 227,000 11 281 302 39 269 1,647 20,245 181,636 441 118 6857 118
pretest (3)
Injected water MLC 1 17.5 8 1200 500 0.88 0.9 0.05 1.2 0.4 110 160 0.06 215 20 0.03
fractest (3)
Recovered water 1.222 50 5.5–5.7 228,000 111 340 384 128 359 2,386 31,606 232,109 295 287 8946 167
fractest
(measured) (4)
Recovered water 1.227 (–) (–) (–) (–) 382 452 151 399 2,625 34,784 272,893 311 301 9906 184
fractest
(calculated)
Horstberg Z1 1.17 40 5.5 n.a. n.a. 260 281 59 302 945 29,582 171,825 92 121 4881 124
produced water
(sample from
2008) (4)
Mg Mn Na NH4 NO2 NO3 O2 PO4 Pb Rb SiO2 SO4 Sr TIC Zn
Injected water 4.0 0.01 24 n.a. 0.004 2.7 9.7 0.04 n.a. n.a. 11 83 0.2 27 0.02
pretest (2)
Recovered water 667 248 86,738 n.a. n.a. n.a. n.a. n.a. n.a. n.a. 37 646 1668 17 646
pretest (3)
Injected water MLC 11 0.08 105 0.68 2.2 9.3 9.3 0.08 n.a. n.a. 9.2 170 1.5 53 0.05
fractest (3)
Recovered water 1021 548 102,238 118 n.a. n.a. n.a. n.a. 228 52 56 76 2512 55 1079
fractest
(measured) (4)
Recovered water 1131 608 123,780 123 n.a. n.a. n.a. n.a. 261 58 59 178 2806 57 1191
fractest
(calculated)
Horstberg Z1 1384 190 69,084 n.a. n.a. n.a. n.a. n.a. 81 n.a. 73 1026 1554 n.a. n.a.
produced water
(sample from
2008) (4)

be expected with CO2 saturation of the recovered water at atmo-
spheric pressure. The injected water had a pH value of 8.0. This va-
lue would decrease with an increase of temperature. Dissolution of
halite or anhydrite would further decrease the pH value, but not
lower than pH 7 at 80 °C. A precipitation of calcite would cause a
further decrease to pH values of ±6.
3.6. Saturation state of the recovered water
For a first orientation the thermodynamical state of mineral
saturation of the recovered water of the fractest was estimated
using the software PhreeqC employing the Pitzer database. As ex-
pected, the water is oversaturated with respect to halite at surface

Fig. 6. d18O and d2H of the injected and recovered water of the fractest. The global
meteoric water line (GMWL) is given for reference (Craig, 1961). The dashed line
indicates mixing curve between injected and hypothetical formation brine.
(VSMOW = Vienna Standard Mean Ocean Water). The star indicates the fresh water
percentage of 40% in the recovered water, calculated with tritium data.
Fig. 7. Solubility of NaCl in pure water as a function of pressure and temperature.
Lines are calculated using data by Pabalan and Pitzer (1987) and Li and Duan
(2011). Open squares indicate GeneSys relevant pressure and temperature
conditions.
134 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139
conditions. In addition, barium and strontium sulphate as well as
calcite and dolomite also show oversaturation.
Literature data show that the maximum halite solubility lies in
the range of the reservoir conditions (170 °C and 600 bar) of
1 1
7.6 mol kgw (442 g kgw ) (Li and Duan, 2011; Pabalan and Pitzer,
1
1987) see Fig. 7, this would lead to a precipitation of 1.4 mol kgw
1
(82 g kgw ) due to the temperature decrease to below 50 °C and
pressure decrease to 1 bar. The temperature dependence of the
precipitation process is more pronounced than the pressure depen-
dence, i.e., the solubility decreases by approx. 4% with an isother-
mal pressure decrease from 800 bar to 1 bar and by approx. 24%
with an isobaric temperature decrease from 170 °C to 20 °C (Li
and Duan, 2011; Pabalan and Pitzer, 1987). Given a cooling to
70 °C and a pressure reduction to 100 bar, as envisaged for the geo-
1 1
thermal usage, a precipitation of 1.1 mol kgw (65 g kgw ) must be
expected.
For barite the maximum solubility at almost reservoir condi-
tions (170 °C, 500 bar and 6 molal NaCl solution) is according to lit-
1 1
erature data 0.8 mmol kgw (187 mg kgw BaSO4) (Blount, 1977).
The calculated barium concentration lies in the same range at
1
1.1 mmol kgw (Table 3). At surface conditions (25 °C, 1 bar) only
1 1
0.17 mmol kgw (40 mg kgw ) of barite (Monnin and Galinier,
1988) is soluble in a 6 molal NaCl solution. Hence a precipitation
1 1
potential of almost 1 mmol kgw (233.4 mg kgw BaSO4), would be
1
expected. The measured sulphate concentration of 0.8 mmol kgw
is less or equal to the combined maximum solubility of barite,
celestite and anhydrite at reservoir conditions (Blount and Dick-
son, 1969; Blount, 1977; Monnin, 1999). At surface conditions
the measured sulphate concentration allows precipitation of both
barite and celestite as found in the salt plug, but not anhydrite.
Because of the elevated iron and manganese concentrations a
high potential for iron and manganese oxide precipitation (up to
1
5.6 mmol kgw ) exists, provided that the water comes into contact
with oxygen.
Fig. 8. SEM pictures of salt plug material. The upper left picture shows the sub-rounded habitus of the halite crystals. The upper right picture shows halite in the background
with (a) native lead and (b) sylvite crystals. The lower picture shows (c) barium–strontium sulphates.
3.7. Salt plug
The salt plug surface in 655 m depth consisted mainly of
mm-sized halite crystals with traces of sylvite and calcium chlo-
ride (CaCl2). Traces of barium-strontium sulphates, native lead
and laurionite (PbCl(OH)) were detected. The halite crystals show
well developed euhedral crystal faces (Fig. 8), but an overall sub-
rounded appearance. The diamond shaped crystallites are sylvite
and the little rods are barium–strontium sulphates. The irregularly
shaped crystallites are native lead. With an overall thickness of al-
most 700 m and if the plug was massive over the whole thickness
the salt plug had have a mass of approx. 30 t.
As expected, the salt plug was enriched in radionuclides com-
pared to the water (Table 4). The radium and lead (210Pb) isotopes
are bound in barium-strontium sulphates and lead phases, respec-
tively. The mineralogical results were in good agreement with the
findings for the radionuclides (Table 2).
4. Discussion
4.1. Origin of salinity
Both recovery tests resulted in the production of saline water,
as it was expected for deep formations in the NGB (Müller and Pap-
endieck, 1975; Wolfgramm et al., 2011). Nevertheless both litera-
ture data and the previous experiences from Horstberg Z1
showed no indication that the recovered water would be oversat-
urated with respect to halite at production conditions. In principle
three explanations for the origin of salinity are possible: (I) the in-
jected fresh water dissolved salt minerals, (II) production of, or
mixing with, salt saturated formation brine or (III) a combination
of both processes. Salt mineral dissolution could be attributed
either to direct contact of the fracture to a rock salt formation
 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139 135

Table 4
Activities of naturally occurring radioactive nuclides in the recovered water of the fractest, in the salt plug (both measured by IRS) and of formation brine of Horstberg Z1
(measured by VKTA Rossendorf e.V.).
226 210 228 228 224 40
Sample Ra SD Pb SD Ra SD Th SD Ra SD K SD
(Bq kg1) (±%) (Bq kg1) (±%) (Bq kg1) (±%) (Bq kg1) (±%) (Bq kg1) (±%) (Bq kg1) (±%)
Recovered water fractest 33.1 6.6 197 18 31.2 10 2.8 11 n.a. 213.3 12
(measured) (4)
Salt plug 2404.2 3.2 1.902.2 12 2.324.8 5 292.3 5.4 n.a. 218.1 24
Horstberg Z1 produced water 22.1 3.3 26.5 7.0 23.7 3.5 0.8 10 33.9 20 112.7 6.6
(2008)

Fig. 9. Comparison of the chemical composition (main ions) of recovered water of pretest, fractest and formation brine of Horstberg Z1, normalised to the recovered water of
the fractest.

(Fig. 2) or to dissolution of dispersed salt cement in the fractured The high salinity indicates that the formation water is close to
formation itself (i.e., Middle Buntsandstein). or in equilibrium with halite. The elution experiments showed,
Generally speaking, the chemical compositions of the recovered that the reservoir rocks contain halite (up to 0.7 wt%). The tritium
water from both tests were very similar (Table 3 and Fig. 9). Recov- data (Table 2) indicate that a considerable percentage (40%) of
ered water from the pretest showed approximately 70% of the fresh water was recovered. As tritium has a half-life time of
salinity encountered in the fractest. The water differed in the con- 12.32 years (Lucas and Unterweger, 2000), the Triassic formation
centration of hydrogen carbonate and redox sensitive elements, brine cannot contain any tritium. Therefore we must assume salt
e.g., iron and manganese and sulphate (Fig. 9). This could be due dissolution in the vicinity of the perforation. Anhydrite hydration
to different enclosure (reaction) times and a higher surface area would cause water consumption and thereby an increase of
in the fractest. The chemical composition of the produced water
at Horstberg Z1 is quite similar to Groß Buchholz Gt1 albeit being
lower in salinity.
The recovered water of both tests had the chemical signature of
a relic evaporation solution, as shown in Fig. 10. The concentration
of bromide compared to the concentrations of sodium and chloride
of both recovered water is too high to originate from halite disso-
lution. The deviation of several ions from the evaporation path can
be attributed to hot water rock interaction (Fig. 11). The strong rel-
ative enrichment of Ca, Sr, Li and Rb in connection with a strong
relative depletion of sulphate and Mg can be explained neither
by evaporation nor by simple mixing of fresh water with a relic
evaporation solution. It is clear, that the original formation brine
must have undergone several processes of hot water rock interac-
tion like weathering of feldspars (Appelo and Postma, 2005), albi-
tisation of plagioclases (Putnis et al., 2008; Schmidt-Mumm,
2008) and dolomitisation of carbonates (Krauskopf and Bird,
1997; Okrusch and Matthes, 2005) which profoundly change the
brine composition over geologic time.
The calculations of the stable isotopes of the formation brine
and the mixing calculations show likewise, that a mixing between Fig. 10. Molar Na/Br and Cl/Br ratios of GeneSys fractest water compared to
fresh and formation water took place. This is indicated by the inter- Na–Cl–Br systematic after Kesler et al. (1995), Fontes et al. (1984) and
section of the mixing line and the lightly shaded box in Fig. 6. McCaffrey et al. (1987).
136 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139

Fig. 11. Comparison of the chemical composition (main ions) of seawater at different evaporation stages of seawater (Fontes et al., 1984) compared to recovered water of the
fractest (normalised to seawater).

Fig. 12. Process visualisation: (A) Injection of fresh water for fracture generation. (B) Sketch perpendicular to sketch A with a reduced scale by a factor of 2.2. Fracture plane
under the assumption of a single fracture. (C) Fracture propagation during fresh water injection. (D) Stop of fresh water injection; fracture convergence; migration of fresh
water into the formation; dissolution of soluble minerals; mixing with formation brine.

salinity in the water. But since gypsum and bassanite are not stable
at reservoir conditions (Blount and Dickson, 1969; Hill, 1937; Hol-
ser, 1979; Ostroff, 1964) we do not expect this process to be of
much importance.
During the pretest, only a small volume of fresh water was
injected into the formation. The water cannot have penetrated
significantly into the formation. A contact to evaporite layers inter-
bedded within the Upper Buntsandstein formation (Röt) which are
located at least 200 m higher above the injection point (Fig. 2) can
therefore be ruled out. The pretest shows that the origin of salt has
to be very close to the perforation, within the target formation
itself.
Taken the dimensions of the fractest, the rock salt mentioned
above could have theoretically become accessible to water circu-
lating along the fracture. Yet, the elevated calcium concentrations
rule out a pure halite source for the salinity encountered. Diffu-
sion-controlled salt transport can be ruled out considering the dis-
tances of almost 200 m to the rock salt horizons and the time
available. Downward transport of dissolved salt via a convection
cell along the fracture plane could be yet another process to gener-
ate salinity. The pretest on the other hand which definitely not
reached the Röt salt, renders this hypothesis rather improbable.
Lower Zechstein salts as a source of salinity are highly unlikely,
as the chemical composition, in particular the magnesium concen-
tration, of the recovered water does not correspond to Zechstein
formation water described in literature (Müller and Papendieck,
1975; Stadler et al., 2012). Additionally no viable upward transport
mechanism can be envisaged.
In conclusion, it is shown that the highly saline water can be
neither a result of halite dissolution from the rock matrix alone,
nor of the recovery of a pure formation brine. The recovered water
of both tests is a mixture which bears a signature of several
processes.
4.2. Process conception
The most likely processes which caused the salinity are concep-
tualised in Fig. 12. The fresh water was injected into the fracture
and into small parts of the surrounding formation and was heated
relatively fast. While injecting the water, the fracture propagated
continuously. After the injection was stopped, fracture propagation
continued. The fracture width decreased to almost zero while the
water migrated a few centimetres into the matrix perpendicular
to the fracture planes (Gulrajani and Nolte, 2000). This results from
 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139
the gradient between the pressure in the fracture and in the forma-
tion. The intruding water dissolved soluble minerals and mixed
with formation brine.
4.3. Isotope exchange at reservoir conditions
The calculations presented above (Section 3.3) considered no
exchange reactions in the reservoir. Since the elevated temperature
of the reservoir would accelerate chemical processes, the temper-
ature dependent exchange reactions of oxygen, hydrogen and sul-
phur should be taken into account. Baertschi (1953) showed strong
increases in exchange rates of, for example oxygen isotopes, at
increasing temperatures. The isotopic composition of oxygen in
the recovered water could be affected by exchange reactions with
clay minerals and dissolved sulphates. The isotopic composition of
hydrogen could be affected by reactions with clay minerals and for
sulphur an exchange with sulphate minerals could be expected.
Oxygen isotope exchange rates between water and sulphates
strongly depend on temperature and pH (Chiba and Sakai, 1985).
For an in situ pH of 4, a half-life of oxygen isotope exchange of four
days at 200 °C was determined. The rate at 100 °C was only given
for pH values lower than 4, for pH 3 they suggest a half-life of
11 years. The oxygen isotope ratio of reservoir anhydrite cement
from the Middle Buntsandstein is more than 20‰ higher than that
of the injected water. For 170 °C and almost six months of ex-
change time, the water would have become enriched in oxygen-18.
An exchange of hydrogen and oxygen could occur with hydro-
xyl ions at the edges of the octahedral layers of clay minerals.
O’Neil and Kharaka (1976) show that water reacting with minerals
becomes isotopically lighter while the minerals are enriched in
heavier isotopes. Unfortunately, the stable isotope composition of
clay minerals in the geothermal reservoir Groß Buchholz Gt1 is un-
known and could cover a very wide range (Hoefs, 2009), an estima-
tion of exchange processes would thus be imprecise. Nevertheless
the exchange should presumably cause depletion of deuterium, as
well as of heavy oxygen, in the water (Liu and Epstein, 1984) and
no enrichment, as it was found. If so, the enrichment of heavier iso-
topes in the recovered water must have been caused by mixing
with isotopically heavier formation brine.
Because tritium is the heaviest of hydrogen isotopes and there-
fore slowest in reaction (Hoefs, 2009), no significant isotope ex-
change processes in the reservoir are to be expected. It would
preferentially remain in the aqueous phase. For that reason tritium
is the most conservative tracer considered here and therefore the
mixing percentages derived are the most significant.
The exchange rate of dissolved sulphate and sulphur minerals is
in general much slower than that of oxygen (Chiba and Sakai,
1985; Ohmoto and Lasaga, 1982). It depends strongly on tempera-
ture and pH. Ohmoto and Lasaga (1982) found the exchange rates
to decrease with increasing pH. At temperatures in the range of
150–200 °C the half-life for sulphur isotope exchange of aqueous
sulphates and sulphides at pH of 4–7 ranged between 4000 years
(150 °C) and 90 years (200 °C).
The in situ pH value was estimated to range around 4 for the
formation brine. The isotope exchange of d34S of dissolved sul-
phates would have a half-life exceeding the six months of enclo-
sure by far (Ohmoto and Lasaga, 1982).
Because no sulphide minerals were found in the samples of the
investigated reservoir rocks, an influence on the overall sulphur
isotope composition can be neglected.
4.4. Dissolution of minerals
The injected oxic fresh water came into contact with several
soluble minerals. These include anhydrite, calcite, dolomite, iron
oxides and halite.
137
If we consider only halite dissolution the measured TDS of
1
452 g kgw (Table 3) for 20,000 m3 of fresh water would correspond
to 7640 t of mobilised salt. It is not necessary for the fresh water to
infiltrate deeply perpendicular to the fracture. The water had a
large interaction area with the reservoir, i.e., a fracture plane of
around 0.6 km2. The porosity of the reservoir rocks is low, with
1–2 vol.%. If there was just 0.2 vol.% of salt minerals in the reservoir
rocks, the fresh water would have to infiltrate only approximately
100 cm (40 cm with a salt content of 0.7 wt%). The calculations
thus show that a small content of salt minerals is sufficient to gen-
erate high salinity.
Halite as a cementation mineral in the Middle Buntsandstein
sandstones was also found in other locations throughout the NGB
(Füchtbauer, 1967, 1974). In Lower Saxony, Nollet et al. (2005)
found centimetre thick halite vein fillings in the Solling sandstones.
In Southwest Denmark, halite was found in Upper Buntsandstein
sandstones (Laier and Nielsen, 1989) and offshore of the Nether-
lands, Purvis and Okkerman (1996) found halite cementation in
the Middle Buntsandstein sandstones.
4.5. Properties of the hypothetical formation brine
The hypothetical formation brine would be oversaturated with
respect to halite, barium-strontium sulphates, some lead phases
and, if oxygen is present, to iron oxides under surface conditions.
At reservoir conditions of almost 600 bar and 170 °C the solubility
1 1
of NaCl amounts to 7.6 mol kgw (442 g kgw ) (Li and Duan, 2011;
Pabalan and Pitzer, 1987).
The high concentrations of manganese and iron in the recovered
water indicate the in situ redox conditions to be reducing. Most
likely, the reservoir redox conditions are even more reducing than
measured, indicated by the presence of intermediate meta-stable
sulphur species and the high content of nitrogen in the exsolved
gas (Appelo and Postma, 2005).
Because of the hydrogen content of the exsolved gas during
both production tests the corrosion of steel casings has to be con-
sidered as an additional factor.
5. Conclusions
The successful fractest at the geothermal well Groß Buchholz
Gt1, Hanover, demonstrates the feasibility of fracture generation
in tight sediments without fracture propping agents. The bottom
temperatures of almost 170 °C are higher than expected. The leach-
ing experiments indicate that dispersed salt is present as a cement
in the Middle Buntsandstein formation in amounts of up to
0.7 wt%. Therefore we assume dissolution of halite minerals in
the reservoir, as well as mixing with formation brine. Due to the
oversaturation of the recovered water with respect to halite at pro-
duction conditions a cooling induced clogging formed a salt plug
between 655 and 1350 m depth.
High salt concentration in deep geothermal reservoirs, as ob-
served here, could also occur at other locations in the NGB within
the Middle Buntsandstein (Nollet et al., 2005). While very saline
yet undersaturated formation brines, as found at Horstberg Z1
and other wells in the NGB (Müller and Papendieck, 1975; Wolf-
gramm et al., 2011) are well known, the dissolution of dispersed
salts is an additional important factor to consider, especially when
generating fractures via fresh water injection in the targeted
formation.
Good understanding of the geochemical environment in the
generated heat exchanger or storage formation is of fundamental
importance for the later geothermal use. Comparing brines of
surrounding wells only gives a first estimate, because the reservoir
formation Buntsandstein is not homogeneous. It varies in its
138 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139
physical and chemical properties over the NGB (Röhling, 2012). If
the formation brine in the reservoir horizon is saturated at under-
ground conditions, salt precipitation has to be expected during
geothermal usage of any kind (Fig. 12).
Therefore specific investigations of core samples of the reservoir
rock over the entire geothermal horizon are strongly encouraged,
especially when exploring new geothermal fields. Preparation of
rock samples for thin section should be done without water.
Elution tests and analysis for sodium and chloride should be
performed with sub cores, uncontaminated by drilling fluid.
A possible improvement for this geothermal project could be a
co-injection of fresh water while producing water from the
fracture. This would help to decrease the saturation of the pro-
duced water. The amount of fresh water, though, must be limited
to prevent uneconomical temperatures. Repeated fresh water
injections to remove salt from the vicinity of the fracture would
be another possible improvement, thereby obtaining subsaturation
with respect to halite for the subsequent production cycles.
Acknowledgments
The authors would like to thank the laboratory staff of the Fed-
eral Institute for Geosciences and Natural Resources, as well as the
team of the GeneSys project. We are grateful to Christian Seeger,
Stefan Schlömer and Michael Schramm for helpful discussions
and technical support. The authors thank the German Federal
Ministry of Economics and Technology (BMWi) for financial
support. We gratefully acknowledge the helpful suggestions of
two anonymous reviewers.
References
Appelo, C.A.J., Postma, D., 2005. Geochemistry, Groundwater and Pollution.
Balkema.
Baertschi, P., 1953. Über die relativen Unterschiede im H218O-Gehalt natürlicher
Wässer. Helvetica Chimica Acta 36 (6), 1352–1369.
Baldschuhn, R., Binot, F., Fleig, S., Kockel, F., 2001. Geotektonischer Atlas von
Nordwest-Deutschland und dem deutschen Nordsee-Sektor - Strukturen,
Strukturentwicklung, Paläogeographie. Geologisches Jahrbuch Reihe A 153, 88.
Blount, W.C., 1977. Barite solubilities and thermodynamic quantities up to 300 °C
and 1400 bars. American Mineralogist 62, 942–957.
Blount, C.W., Dickson, F.W., 1969. The solubility of anhydrite (CaSO4) in NaCl–H2O
from 100 to 450 °C and 1 to 1000 bars. Geochimica et Cosmochimica Acta 33,
227–245.
Chiba, H., Sakai, H., 1985. Oxygen isotope exchange rate between dissolved sulfate
and water at hydrothermal temperatures. Geochimica et Cosmochimica Acta
49, 993–1000.
Chiba, H., Kusakabe, M., Hirano, S.-I., Matsuo, S., Somiya, S., 1981. Oxygen isotopic
fractionation factors between anhydrite and water from 100 to 550 °C. Earth
and Planetary Science Letters 53, 55–62.
Clark, I.D., Fritz, P., 1997. Environmental Isotopes in Hydrogeology. CRC Press LLC.
Craig, H., 1961. Isotopic variation in meteoric waters. Science 133, 1702–1703.
Fontes, J.C., Brissaud, I., Michelot, J.L., 1984. Hydrological implications of deep
production of chloride-36. Nuclear Instruments and Methods in Physics
Research B5, 303–307.
Füchtbauer, H., 1967. Der Einfluss des Ablagerungsmilieus auf die
Sandsteindiagenese im Mittleren Buntsandstein. Sedimentary Geology 1, 159–
179.
Füchtbauer, H., 1974. Zur Diagenese fluviatiler Sandsteine. Geologische Rundschau
63 (3), 904–925.
Gilg, H.A., Sheppard, S.M.F., 1996. Hydrogen isotope fractionation between kaolinite
and water revisited. Geochimica et Cosmochimica Acta 60 (3), 529–533.
Gulrajani, S.N., Nolte, K.G., 2000. Reservoir Stimulation. In: Economides, M.J., Nolte,
K.G. (Eds.), third ed. third ed. John Wiley & Sons, Inc., pp. 1–63.
Heinig, S., Hesshaus, A., Röhling, H.-G., 2011. Geothermie-Projekt GeneSys
Hannover: Sedimentpetrographische und geochemische Untersuchungen
geringpermeabler Sedimente des Mittleren Buntsandsteins.
Hill, A.E., 1937. The transition temperature of gypsum to anhydrite. Contribution
From the Chemical Laboratorium of New York University 59, 2242–2244.
Hoefs, J., 2009. Stable Isotope Geochemistry, fifth ed. Springer Verlag.
Holser, W.T., 1979. Marine Minerals. In: Burns, R.G. (Ed.). Mineralogical Society of
America, pp. 235–275.
Kehrer, P., Orzol, J., Jung, R., Jatho, R., 2005. Erschließungskonzepte zur
Erdwärmenutzung aus gering-permeablen Sedimentgesteinen. Erdöl Erdgas
Kohle 121 (4), 151–155.
Kesler, S.E., Appold, M.S., Martini, A.M., Walter, L.M., Huston, T.J., Kyle, J.R., 1995. Na-
Cl-Br systematics of mineralizing brines in Mississippi Valley–type deposits.
Geology 23 (7), 641–644.
Krauskopf, K.B., Bird, D.K., 1997. Introduction to Geochemistry, third ed. McGraw-
Hill, New York.
Laier, T., Nielsen, B.L., 1989. Cementing halite in Triassic Bunter Sandstone (Tönder,
southwest Denmark) as a result of hyperfiltration of brines. Chemical Geology
76, 353–363.
Li, J., Duan, Z., 2011. A thermodynamic model for the prediction of phase equilibria
and speciation in the H2O–CO2–NaCl–CaCO3–CaSO4 system from 0 to 250 °C, 1
to 1000 bar with NaCl concentrations up to halite saturation. Geochimica et
Cosmochimica Acta 75, 4351–4376.
Liu, K.-K., Epstein, S., 1984. The hydrogen isotope fractionation between kaolinite
and water. Isotope Geoscience 2, 335–350.
Lloyd, R.M., 1968. Oxygen isotope behavior in the sulfate-water system. Journal of
Geophysical Research 73 (18), 6099.
Lucas, L.L., Unterweger, M.P., 2000. Comprehensive review and critical evaluation of
the half-life of tritium. Journal of Research of the National Institute of Standards
and Technology 105 (4), 541–549.
McCaffrey, M.A., Lazlar, B., Holland, H.D., 1987. The evaporation path of seawater
and the coprecipation of Br- and K+ with halite. Journal of Sedimentary
Petrology 57 (5), 928–937.
Mizutani, Y., Rafter, A., 1969. Oxygen isotopic composition of sulphates. Part 3,
oxygen isotopic fractionation in the bisulphate ion-water system. New Zealand
Journal of Science 12, 54–59.
Monnin, C., 1999. A thermodynamic model for the solubility of barite and celestite
in electrolyte solutions and seawater to 200 °C and to 1 kbar. Chemical Geology
153, 187–209.
Monnin, C., Galinier, C., 1988. The solubility of celestite and barite in electrolyte
solutions and natural waters at 25 °C: a thermodynamic study. Chemical
Geology 71, 283–296.
Müller, E.P., Papendieck, G., 1975. Zur Verteilung, Genese und Dynamik von
Tiefenwässern unter besonderer Berücksichtigung des Zechsteins. Zeitung der
geologischen Wissenschaft Berlin 3, 167–196.
Nollet, S., Hilgers, C., Urai, J., 2005. Sealing of fluid pathways in overpressure cells: a
case study from the Buntsandstein in the Lower Saxony Basin (NW Germany).
International Journal of Earth Sciences (Geology Rundsch) 94, 1039–1055.
Ohmoto, H., Lasaga, A.C., 1982. Kinetics of reaetions between aqueous sulfates and
sulfides in hydrothermal systems. Geochimica et Cosmochimica Acta 46, 1727–
1745.
Okrusch, M., Matthes, S., 2005. Mineralogie – Eine Einführung in die spezielle
Mineralogie, Petrologie und Lagerstättenkunde. Springer Verlag.
O’Neil, J.R., Kharaka, Y.K., 1976. Hydrogen and oxygen isotope exchange reactions
between clay minerals and water. Geochimica et Cosmochimica Acta 40, 241–
246.
O’Neil, J.R., Clayton, R.N., Mayeda, T.K., 1969. Oxygen isotope fractionation
in divalent metal carbonates. Journal of Chemical Physics 51 (12), 5547–
5558.
Ostroff, A.G., 1964. Conversion of gypsum to anhydrite in aqueous salt solutions.
Geochimica et Cosmochimica Acta 28, 1363–1372.
Pabalan, R.T., Pitzer, K.S., 1987. Thermodynamics of concentrated electrolyte
mixtures and the prediction of mineral solubilities to high temperatures for
mixtures in the system Na–K–Mg–Cl–SO4–OH–H2O. Geochimica et
Cosmochimica Acta 51, 2429–2443.
Parkhurst, D.L., Appelo, C.A.J., 1999. User’s guide to PHREEQC (Version 2) – a
computer program for speciation, batch-reaction, one-dimensional transport,
and inverse geochemical calculations. In: Water-Resources Investigations,
Report 99–4259.
Pitzer, K.S., 1981. Characteristics of very concentrated aqueous solutions. Physics
and Chemistry of the Earth 13 (14), 249–272.
Purvis, K., Okkerman, J.A., 1996. Geology of Gas and Oil under the Netherlands. In:
Rondeel, H.E., Batjes, D.A.J., Nieuwenhuijs, W.H. (Eds.). Kluwer Academic
Publishers, pp. 179–189.
Putnis, A., Engvik, A., Gerald, J.F., Hövelmann, J., 2008. Albitisation of plagioclase in
nature and experiment. In: International Geological Congress Oslo.
Röhling, H.-G., 2012. Der Buntsandstein im Norddeutschen Becken. Deutsche
Stratigraphische Kommission.
Röhling, H.-G., Heinig, S., 2012. Lithostratigraphie und Petrographie des Mittleren
Buntsandsteins in der Geothermiebohrung Groß Buchholz Gt 1 und der
Bohrung Hämelerwald Z1.
Schäfer, F., Hesshaus, A., Hunze, S., Jatho, R., Luppold, F.-W., Orilski, J., Plaetsch, T.,
Röhling, H.-G., Tischner, T., Wonik, T., 2012. Kurzprofil der Geothermiebohrung
Groß Buchholz Gt1. Erdöl Erdgas Kohle 128 (1), 20–26.
Schmidt-Mumm, A., 2008. Albitisation – from mineral grains to giant ore deposits.
In: International Geological Congress Oslo.
Schröder, H., Hesshaus, A., 2009. Langfristige Betriebssicherheit geothermischer
Anlagen – BGR-Abschlussbericht.
Schulz, R., Magemar, T., Alten, J.-A., Kühne, K., Maul, A.-A., Pester, S., Wirth, W., 2007.
Aufbau eines geothermischen Informationssystems für Deutschland. Erdöl
Erdgas Kohle 123 (2), 76–81.
Seeger, C., 2011. Projektierung einer Seperationsanlage zur physikalischen
Trennung einer Gasphase aus Thermalwasser.
Stadler, S., Sültenfuß, J., Holländer, H.M., Bohn, A., Jahnke, C., Suckow, A., 2012.
Isotopic and geochemical indicators for groundwater flow and multi-
component mixing near disturbed salt anticlines. Chemical Geology 294–295,
226–242.
 A. Hesshaus et al. / Physics and Chemistry of the Earth 64 (2013) 127–139 139

Tischner, T., Evers, H., Hauswirth, H., Jatho, R., Kosinowski, M., Sulzbacher, H., 2010. New
Concepts for Extracting Geothermal Energy from one Well: The GeneSys-Project.
Tischner, T., Krug, S., Pechan, E., Hesshaus, A., Jatho, R., Bischoff, M., Wonik, T., 2013.
Massive hydraulic fracturing in low permeable sedimentary rock in the GeneSys
project. In: Proceedings, Thirty-Eighth Workshop on Geothermal Reservoir
Engineering, Stanford University, Stanford, California, February 11–13, 2013.
Wolfgramm, M., Thorwart, K., Rauppach, K., Brandes, J., 2011. Zusammensetzung,
Herkunft und Genese geothermaler Tiefengrundwässer im Norddeutschen
Becken (NDB) und deren Relevanz für die geothermische Nutzung. Zeitschrift
Fur Geologische Wissenschaften 39 (3/4), 173–193.