14 C dating of Gorleben groundwater G. Buckau a, *, R. Artinger a , S. Geyer b , M. Wolf c , P. Fritz b , J.I. Kim a a Institut fu ¨r Nukleare Entsorgungstechnik, Forschungszentrum Karlsruhe, P.O. Box 3640, D-76021 Karlsruhe, Germany b UfZ-Umweltforschungszentrum Leipzig-Halle, D 06120 Halle, Germany c GSF-National Research Center for Environment and Health, Institute of Hydrology, D-85764 Neuherberg, Germany Received 19 July 1998; accepted 31 May 1999 Editorial handling by W.M. Edmunds Abstract Previous attempts to apply 14 C for dating of groundwater in the Gorleben aquifer system has given results with conflicting 3 H and stable isotope data and hydrological estimates. 14 C model ages of 1–10 ka have been found for 3 H containing recharge water, up to 31 ka for groundwater with Holocene stable isotope signatures and 6–10 ka for groundwater at 35 m depth. In this paper it is shown, that the reasons are assumption of to high 14 C concentration in recharge groundwater and not correcting for the influence of 14 C dilution by dissolved inorganic C (DIC) from microbiologically mediated mineralization of organic components in deep sediments. To overcome these diculties a new approach is applied evaluating the site-specific 14 C source term (including the influence of nuclear atmospheric testing), and the already previously used overall dilution of DIC. Closed system conditions are assumed and no isotopic fractionation is considered. For most of the groundwaters, the 14 C ages achieved by the present method are in agreement with 3 H, stable isotopes and hydrological estimates. It is shown that down to approximately 140 m depth no 14 C decay can be detected. Situations are also discussed, either where the 14 C method is not applicable (shallow peat-bog like groundwater) or does not yield reliable groundwater ages (brines at <200 m depth). # 2000 Elsevier Science Ltd. All rights reserved. 1. Introduction Within the German program for nuclear waste dis- posal, the Gorleben salt dome has been investigated as a candidate site for more than 20 a. To ensure long- term safety, amongst others, the radiological impact from a release of long-lived radionuclides into the overlieing aquifer system needs to be assessed. Thereby, adequate understanding of the hydrological situation is required. For this purpose of hydrological modeling, groundwater dating is essential. Despite large eorts, there is still a great deal of uncertainty about the origin and age of the Gorleben ground- waters. Groundwater dating by stable isotopes, Noble gas temperatures and 14 C lead to inconclusive or con- tradicting results. The aim of this paper is to show that, contrary to previous experiences, 14 C ground- water dating can be performed on many of the groundwaters. It is also the aim to show where this is not the case due to lack in understanding of origin and geochemical conditions. Application of the 14 C groundwater dating relies on dissolved inorganic C (DIC) as a groundwater tracer. The age is calculated based on the dierence between the 14 C concentration of DIC in a given groundwater Applied Geochemistry 15 (2000) 583–597 0883-2927/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0883-2927(99)00071-2 * Corresponding author. Tel.: +49-7247-824-461; fax: +49- 7247-826-998. E-mail address: [email protected] (G. Buckau).
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
14C dating of Gorleben groundwater
G. Buckaua,*, R. Artingera, S. Geyerb, M. Wolfc, P. Fritzb, J.I. Kima
cGSF-National Research Center for Environment and Health, Institute of Hydrology, D-85764 Neuherberg, Germany
Received 19 July 1998; accepted 31 May 1999
Editorial handling by W.M. Edmunds
Abstract
Previous attempts to apply 14C for dating of groundwater in the Gorleben aquifer system has given results withcon¯icting 3H and stable isotope data and hydrological estimates. 14C model ages of 1±10 ka have been found for3H containing recharge water, up to 31 ka for groundwater with Holocene stable isotope signatures and 6±10 ka for
groundwater at 35 m depth. In this paper it is shown, that the reasons are assumption of to high 14C concentrationin recharge groundwater and not correcting for the in¯uence of 14C dilution by dissolved inorganic C (DIC) frommicrobiologically mediated mineralization of organic components in deep sediments. To overcome these di�culties a
new approach is applied evaluating the site-speci®c 14C source term (including the in¯uence of nuclear atmospherictesting), and the already previously used overall dilution of DIC. Closed system conditions are assumed and noisotopic fractionation is considered. For most of the groundwaters, the 14C ages achieved by the present method are
in agreement with 3H, stable isotopes and hydrological estimates. It is shown that down to approximately 140 mdepth no 14C decay can be detected. Situations are also discussed, either where the 14C method is not applicable(shallow peat-bog like groundwater) or does not yield reliable groundwater ages (brines at <200 m depth). # 2000Elsevier Science Ltd. All rights reserved.
1. Introduction
Within the German program for nuclear waste dis-posal, the Gorleben salt dome has been investigated asa candidate site for more than 20 a. To ensure long-
term safety, amongst others, the radiological impactfrom a release of long-lived radionuclides into theoverlieing aquifer system needs to be assessed.
Thereby, adequate understanding of the hydrologicalsituation is required. For this purpose of hydrological
modeling, groundwater dating is essential. Despitelarge e�orts, there is still a great deal of uncertaintyabout the origin and age of the Gorleben ground-
waters. Groundwater dating by stable isotopes, Noblegas temperatures and 14C lead to inconclusive or con-tradicting results. The aim of this paper is to showthat, contrary to previous experiences, 14C ground-
water dating can be performed on many of thegroundwaters. It is also the aim to show where this isnot the case due to lack in understanding of origin and
geochemical conditions.Application of the 14C groundwater dating relies on
dissolved inorganic C (DIC) as a groundwater tracer.
The age is calculated based on the di�erence betweenthe 14C concentration of DIC in a given groundwater
Applied Geochemistry 15 (2000) 583±597
0883-2927/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
and its concentration at the time of the recharge (the
source term). This approach encounters two majorproblems, namely (i) quanti®cation of the source term,and (ii) decrease of the 14C concentration by mixing
with DIC from various sources (generally with zero14C concentration) due to geochemical reactions in thecourse of the groundwater evolution.
The 14C concentration in DIC upon primaryrecharge is normally set to the level of atmospheric
CO2 prior to nuclear atmospheric testing, i.e. around100 pmc ( percent modern C referred to 94.9 % of theactivity concentration of the NBS oxalic acid standard
in 1950) (Pearson and White, 1967; Pearson and Han-shaw, 1970). In the atmosphere, a deviation of the 14C
concentration from 100 pmc is expected from changesin the earth's magnetic ®eld and activity of the sun.Climatic changes may also in¯uence the 14C concen-
tration in di�erent global reservoirs (atmosphere, bio-sphere and oceans) (Stuiver et al., 1991; Stuiver andBraziunas, 1993). Consequently, the 14C age may devi-
ate from the actual groundwater age (Pearson et al.,1983). The di�erences increase steadily to greater than
3 by 25 ka BP. For groundwater dating, however, suchdi�erences are negligible compared to other uncertain-ties. A much more dramatic impact resulted from the
atmospheric nuclear testing leading to a peak valueapproaching 200 pmc in the northern hemisphere
around 1963±1966 (Moser and Rauert, 1983; Levinand Kromer, 1997). Therefore, it is di�cult to deter-mine the 14C age of the primary recharge of a pre-at-
mospheric nuclear testing groundwater. A directin®ltration of DIC from atmospheric CO2 via precipi-tation is much less than dissolution of soil-CO2, which
originates from root-respiration and turnover of soilorganic material, including humus and peat. The 14C
concentration of primary recharge, i.e. DIC of organicorigin at shallow depth, depends on the 14C concen-trations of such di�erent sources and their contribution
to DIC. The 14C concentration of DIC from root-res-piration will directly re¯ect the atmospheric concen-tration. Through turnover of relatively old humus or
peat, the 14C concentration of primary recharge DIC islowered (Taylor, 1998).
Once the primary recharge DIC is generated, di�er-ent processes may lead to its isotopic modi®cation, es-pecially dilution by various sources. Published
examples on such processes are: (1) dissolution of 14Cfree sedimentary carbonate by the primary recharge
Cic acid (MuÈ nnich, 1957; Ingerson and Pearson, 1964);(2) on repeated dissolution and precipitation, the car-bonate sediments may contain 14C (Geyh, 1972; Fritz
et al., 1978); (3) ion exchange of Ca2+ and Mg2+
against Na+ in clay minerals, leading to dissolution ofCaCO3 (Fritz and Mozeto, 1981); (4) incongruent dis-
solution of dolomite (Matthess, 1990); (5) degassing ofCO2 (Eichinger, 1981); (6) dissolution of CaSO4 result-
ing in calcite precipitation through excess concen-
trations of Ca2+ (Eichinger, 1981); (7) changes insaturation concentrations via weathering of feldspar(Vogel and Ehhalt, 1963); (8) dissolution of sedimen-
tary carbonate by organic acids, such as humic andfulvic acids (Vogel and Ehhalt, 1963); (9) admixing ofvolcanic or magmatic CO2 (Fritz and Mozeto, 1981;
generation (Barker et al., 1979); and (11) microbiologi-
cally mediated oxidation of organic C via dissolved O2
or nitrate (Stumm and Morgan, 1996) or SO4 (Ceder-strom, 1946; Pearson and White, 1967; Pearson and
Hanshaw, 1970). Thereby, DIC and dissolved organicC (DOC) are generated in groundwater (Buckau et al.,
2000).In order to cope with the di�erent sources leading to
modi®cation of the isotopic C composition of DIC,
di�erent models have been developed which considerrelevant chemical processes and di�erences in the 13Csignatures. The simplest approach of chemical mixing
takes account of the dissolution of sedimentary car-bonates by primary recharge DIC (MuÈ nnich, 1957;
Ingerson and Pearson, 1964; Tamers, 1975). Extendedchemical mixing models utilize the chemical speciationby PHREEQE (Parkhurst et al., 1980) modi®ed for
appraisal of the impact of DIC on the 14C content(Wolf and Rhode, 1992; Wolf et al., 1993). The sim-
plest isotopic mixing model utilizes di�erent 13C con-centrations of DIC from primary recharge and theadmixed DIC from carbonate sediments. Such an
approach is applicable only for the two source systemwhen their 13C signatures are known. More complexisotopic mixing models attempt to account for ad-
ditional impacts from isotopic fractionation and isoto-pic exchange (Mook, 1976; Reardon and Fritz, 1978;
Fontes and Garnier, 1979). Some models combine iso-topic mixing with hydrochemical groundwater evol-ution (Wigley et al., 1978; Plummer et al., 1990, 1994).
One source of additional DIC in groundwater isthe mineralization/oxidation of sedimentary organic
C (SOC). The groundwater age indicative primaryrecharge DIC and the 14C free DIC from mineraliz-ation of SOC originate from organic sources and thus
cannot be distinguished from each other by the 13Ccontent. Therefore, in aquifer systems where this pro-cess is of considerable magnitude, radiocarbon dating
based on 13C to quantify the primary recharge DICbecomes falsi®ed. In above models this process is not
accounted for. If su�cient data are available, includingchemical composition of groundwater, isotopic compo-sition of dissolved C containing species and mineralogy
of sediments, the in¯uence of such processes can beaccounted for along a given ¯ow-path (Aravena et al.,1995). In the Gorleben aquifer system adequate miner-
alogical data are missing. Furthermore, detailed knowl-edge of the hydrological situation is not available and
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597584
considerable mixing of groundwater is expected
(Buckau et al., 1999). Therefore, in this paper another
approach is used.
As mentioned above, the aim of the paper is to
develop a method for 14C groundwater dating of
the Gorleben groundwater but also to show where
lack of information makes such dating unreliable.
For this purpose, the impact of mineralization of SOC
is accounted for by regarding the overall dilution of
DIC from recharge to a given groundwater, irrespec-
tive of the DIC source (Ingerson and Pearson, 1964;
Pearson and White, 1967; Pearson and Hanshaw,
1970; Wigley et al., 1978; Pearson, 1991). The
approach furthermore makes use of a site-speci®c 14C
source term, taking into account impact of atmos-
pheric nuclear testing. The result is plausible ground-
water ages instead of the excessive ages found by other
approaches.
2. The Gorleben aquifer system
The Gorleben aquifer system is situated above a Per-mian salt dome at Gorleben in the North German
Plain (Lower Saxony) (Fig. 1). The aquifer systemextends to 280 m below the ground surface in
reworked and deformed Miocene and Pleistocene gla-cial sediments. Below approximately 200 m depth the
so-called Gorleben channel is found with brines indirect contact with the salt dome. The main recharge
area is located south of the salt dome. Signi®cant 3Hcontents <1 TU (i.e. groundwater with ages <40 a)
are only found in groundwater samples down to adepth of 25 m (Suckow, 1993). After sinking to lower
aquifer layers, groundwater is expected to ¯ow mainlytowards the NW. At depths greater than approxi-
mately 200 m, the Gorleben channel becomes a pre-ferred ¯owpath running to the NE. The sediments
Fig. 1. Gorleben site with sampling locations.
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597 585
contain local Miocene brown coal and Pleistocene peat
deposits and groundwater has a�ected the mineraliz-
ation of SOC. By this process, DOC and DIC are sim-
ultaneously generated resulting in enhanced DICconcentrations in conjunction with DOC concen-
trations reaching <200 mg C/L (Fig. 2). The process
is driven by microbiologically mediated SO4 reduction
leading to enhanced DOC concentrations and SO4 de-pletion, especially in groundwaters between 50 and 170
m (Fig. 3). Details can be found in Buckau et al.
(1999, 2000).
In this paper, the groundwaters are classi®ed as (see
Table 1; Artinger et al., 1999): (i) recharge; (ii) tran-
sition; (iii) enhanced DOC; (iv) channel brines; and (v)
uplift-mixing groundwater. Five recharge groundwatersfrom depths <25 m have signi®cant 3H content and
DOC concentrations of 2.2 mg C/L or less. Two tran-
sition groundwaters from depths of around 30 and 70
m have low DOC concentrations similar to therecharge samples, but no signi®cant 3H content. Chan-
nel brines (3 samples, depths <200 m) also have low
DOC concentrations. Two groundwaters at depths of
65±73 m show moderately enhanced DOC concen-trations, and another six groundwaters at 35±137 mdepth have highly enhanced DOC concentrations.These eight groundwaters are called enhanced DOC
groundwaters. The uplift-mixing Gohy-341 is found atshallow depth (10±13 m). Contrary to the aboverecharge groundwaters, however, it has a high DOC
concentration (64 mg C/L). The salt content of thisgroundwater is dominated by NaCl originating fromsalt dome dissolution at great depth. However, this
groundwater also has some 3H and a 14C concen-tration indicative of recent C sources. Therefore, thisgroundwater is considered as a mixture of old deepgroundwater and recent recharge.
3. Methodology
For 14C determination, carbonate-free NaOH andBaCl2. H2O was added to at least 50 l of water col-
lected in polyethylene containers ¯ushed with N2 or Arto exclude atmospheric CO2. The closed containerswere allowed to stand overnight prior to separation of
BaCO3 precipitate under inert-gas. After CO2 gener-ation from BaCO3 and benzene synthesis (Eichinger etal., 1980), 15 to 60 mg Butyl-PDB scintillator (Packard
& Co.) was added to 1 to 4 mL benzene, and 14C wasmeasured by liquid scintillation counting (Quantulus,LKB-Wallac). The experimental detection limit is
Fig. 2. DOC versus DIC concentrations in Gorleben ground-
water. Through mineralization of sedimentary organic C,
DOC (consisting mainly of humic and fulvic acids) and DIC
are in situ generated. The direct relationship between DOC
and DIC falls into two groups; at lower pH additional sedi-
mentary carbonate is dissolved (for details, see Artinger et al.,
1999; Buckau et al., 1999, 2000) (®gure from Buckau et al.,
1999).
Fig. 3. SO4 depletion and in situ generation of DOC in con-
junction with mineralization of sedimentary organic C.
Shaded lines show values/region where SO4 concentrations
are expected from mixing of deep brines and recharge ground-
water. SO4 depletion and enhanced DOC concentrations are
found especially between 50 and 170 m (®gure from Buckau
et al., 1999, where details can be found). Lower limit for sul-
phate reduction from Appelo and Postma (1996).
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597586
Table
1
Sampled
Gorleben
groundwaters,samplingdepths,
pH,DIC
concentration
and
isotopic
characterization
ofgroundwaterand
DIC
.Analyticalerrors
(2s);
2H:21-,
18O:
20.15-
and
13C:2
0.4
-
Sample
No.
Sample
Sampling
depth
(m)
pH
DOC
(mgC/L)
d2H
(-)a
d18O
(-)a
3H (TU)b
22s
DIC
(mmol/L)
d13C
(-)c
14C
(DIC
)
(pmc)
d22s
Rechargearea
1Gohy-181
15±18
6.4
0.5
ÿ62.3
ÿ8.852
0.08
38.7
22.7
0.67
ÿ20.6
45.7
20.9
2Gohy-411
17±20
7.0
2.2
ÿ61.7
ÿ8.422
0.11
20.7
21.5
1.54
ÿ17.9
54.2
23.3
3Gohy-421
10±13
6.0
1.9
ÿ62.2
ÿ8.612
0.06
20.0
21.5
0.84
ÿ23.1
68.5
24.5
4Gohy-611
21±24
8.4
1.4
ÿ62.0
ÿ8.542
0.19
8.0
20.7
1.06
ÿ13.3
26.5
21.7
5Gohy-711
6±9
6.0
0.9
ÿ64.4
ÿ9.082
0.09
29.2
22.1
0.59
ÿ22.4
70.4
21.4
Transitionarea
6Gohy-182
70±73
8.0
0.8
ÿ63.6
ÿ8.902
0.04
<0.28
1.07
ÿ13.4
27.9
20.6
7Gohy-201
30±35
7.8
0.9
ÿ62.3
ÿ8.852
0.28
<1.1
0.79
ÿ13.6
26.5
22.1
EnhancedDOC
8Gohy-412
65±68
7.6
7.6
ÿ57.6
ÿ8.40
<0.23
3.66
ÿ6.8
8.6
20.5
9Gohy-492
35±38
8.0
127
ÿ58.1
ÿ8.462
0.22
<0.17
11.69
ÿ11.1
2.1
21.2
10a
Gohy-572(I)
70±73
8.8
14.4
ÿ58.1
ÿ8.552
0.16
0.502
0.12
4.75
ÿ4.0
4.6
21.8
10b
Gohy-572(II)
70±73
````
ÿ59.8
0.582
0.19
4.75
ÿ4.1
5.2
21.1
11a
Gohy-573(I)
134±137
8.0
97.2
ÿ59.7
ÿ8.622
0.20
<0.13
8.15
ÿ9.1
4.0
20.7
11b
Gohy-573(II)
134±137
````
ÿ61.5
<0.40
8.15
ÿ9.2
3.7
20.8
12
Gohy-612
121±125
8.4
184
ÿ60.4
ÿ8.592
0.24
0.522
0.21
13.34
ÿ11.9
1.9
20.8
13
Gohy-2211
83±85
8.1
93.6
ÿ59.6
ÿ8.622
0.08
0.322
0.13
9.52
ÿ10.0
5.2
20.8
14
Gohy-2227
128±130
7.8
73.4
ÿ62.4
ÿ8.852
0.10
<0.3
8.17
ÿ12.8
5.0
20.8
Channel
brines
15
Gohy-193
220±223
6.8/7.1
e3.2
ÿ72.2
ÿ9.862
0.04
<0.20
4.76
ÿ14.3
1.8
20.3
16
Gohy-514
235±238
6.3/7.0
e1.6
ÿ80.7
ÿ10.312
0.29
<0.25
3.65
ÿ12.8
5.1
21.4
17
Gohy-653
216±219
6.5/7.1
e1.3
ÿ67.7
ÿ9.482
0.07
<0.26
4.39
ÿ12.5
7.6
20.7
Uplift-m
ixing
18
Gohy-341
10±13
7.0
64.0
ÿ58.8
ÿ8.362
0.09
0.412
0.12
5.49
ÿ15.1
25.0
21.8
aRel.V-SMOW
(meanvalues
ofmultiple
samplingbetween1979and1993).
bTU
3H
unit(1
TU=
0.118Bq/L).
cRel.V-PDB.
dPercentmodernC
(94.9%
of14C
activityconcentrationin
NBSoxalicacidstandard
in1950).
epH
correctedforsaltin¯uence
accordingto
Runde(1993).
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597 587
between 0.6 and 0.7 pmc (2s ) (background <0.3 cpm,counting e�ciency approximately 82 %, counting time
1000 min). 3H was determined by liquid scintillationcounting of water or gas counting of propane afterelectrolytic enrichment of 3H by a factor of approxi-
mately 20 (Eichinger et al., 1980; Wolf et al., 1981;Wolf and Singer, 1991). The detection limit for thesetwo methods are 0.7 and 0.1±0.2 TU, respectively. 13C
of DIC and 2H and 18O of the water matrix was deter-mined by mass spectrometry. Bicarbonate wasmeasured by titration and DIC calculated from known
carbonate equilibrium data. More detailed descriptionof experimental procedure can be found in (Kim et al.,1995).
4. Results and discussion
Results in this paper are partly based on unpub-
lished data from `Bundesanstalt fuÈ r Geowissenschaftenund Rohsto�e` (BGR, Hannover, Germany), the `Bun-desamt fuÈ r Strahlenschutz` (BfS, Salzgitter, Germany)and `Institut fuÈ r Radiochemie der Technischen Univer-
sitaÈ t MuÈ nchen` (TU Munich, Germany).Results and discussion are divided into 5 parts: (i)
determination of groundwater age from 3H, 2H and18O data; (ii) comparison of these groundwater ageswith results from published 14C methods; (iii) decreasein the 14C concentration of DIC by dilution through
the mineralization of SOC; (iv) evaluation of thesource term, i.e. concentration and 14C content of pri-marily recharge DIC; and (v) 14C groundwater datingby the present approach.
4.1. Groundwater age ranges
Table 1 shows the results of the isotopic characteriz-ation of the groundwaters. Recharge groundwaterscontain considerable concentrations of 3H (between
approximately 20 and 40 TU) and thus ages of <40 acan be postulated. The 3H concentration of Gohy-611(8 TU) can be attributed to mixing of recent and older
groundwater. In the shallow (10±13 m) uplift-mixinggroundwater Gohy-341 3H is found in the order of 1% of the content in recently recharged groundwaters.Amongst the groundwaters at depth below 30 m, the
transition waters and channel brines have 3H concen-trations below detection limits of the applied methods.Of the enhanced DOC groundwaters, Gohy-572, -612
and -2211 have 3H concentrations between 0.32 and0.54 TU. These 3 samples are from 70±125 m abovethe salt dome. Other enhanced DOC groundwaters in
this area, from 128±137 m depth (Gohy-573 and -2227), have 3H concentrations below detection limit.The same is true for the enhanced DOC groundwaters
Gohy-412 (65±68 m) (from the SW part of therecharge area south of the salt dome) and Gohy-492
(35±38 m) somewhat north of the other enhancedDOC groundwaters (cf. Fig. 1). The good reproducibil-ity of sampling and measurement can be seen from
repeated sampling of Gohy-572 and -573 (Table 1).While the positive 3H indications for 3 of the enhancedDOC samples may indicate the presence of a young
component, it is also possible that they are due to de-®ciencies in the well casing. In summary, for the pur-pose of establishing groundwater age ranges, 3H
concentrations <15 TU are allocated to a ground-water age <40 a, <1.1 TU to <40 a and 8 TU inGohy-611 to 140 a.Another indicator for the groundwater age is the 2H
and 18O content where precipitation will mainly varyalong the global meteoric water line (GMWL(d 2H=8xd18O+10) (Craig, 1961)). Deviation may
occur through, for example fractionation via evapor-ation of surface water. In Fig. 4, the 2H and 18O con-centrations are plotted. For reference to young
recharge groundwaters (<40 a) from comparablerecharge conditions, data from the Fuhrberg aquifersystem are also shown (Kim et al., 1995). Furthermore
data from a broader study with more emphasis ondeep brines are shown (Suckow, 1993). The bulk of thedata fall close to and somewhat below the GMWL.Groundwaters with 18O signatures more positive than
approximately ÿ9.1- can be allocated to Holoceneorigin (<10 ka). Some data referring to groundwaters
Fig. 4. 2H and 18O concentrations in Gorleben groundwaters
and Fuhrberg groundwaters for reference to young recharge
groundwaters from comparable recharge conditions.
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597588
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597 589
with less negative isotope signatures from depths of67±121 m below sea level fall below the GMWL
(upper right part of Fig. 4) possibly indicating an ori-gin with considerable impact from surface reservoirevaporation.
With respect to de®ning the 2H and 18O isotopicrange signifying groundwaters of Pleistocene origin(<12 ka), absolute reference values are missing. From
Fig. 4 it appears reasonable to assume that 18O valuesmore negative than about ÿ11- represent Gorlebengroundwaters of Pleistocene origin with mixed ground-
waters in the intermediate region between ÿ9.1 andÿ11-. On this basis one may expect Gohy-514 to bemainly of Pleistocene origin and Gohy-193 and -653 tobe mixtures of Pleistocene and Holocene origin.
The age ranges suggested by 3H and stable isotopedata are listed in the third column of Table 2. In therecharge area, 3H containing groundwaters at 6±24 m
depth are expected to have ages <40 a, possibly witholder components in Gohy-611. The transition andenhanced DOC groundwaters between 30 and 137 m
are expected to have ages between 40 and 10 ka. Asmentioned above, a meaningful groundwater age esti-mation for the uplift-mixing groundwater Gohy-341
appears questionable. For the channel brines Gohy-193 and -653 the expected ages can be set to the tran-sition between Pleistocene and Holocene (110 ka)whereas Gohy-541 is expected to be of Pleistocene ori-
gin (<12 ka).
4.2. Comparison with the results from published 14C
methods
Groundwater ages calculated using di�erent models/methods are listed in Table 2 (Kim et al., 1995). The14C concentrations are shown in Table 1 and necessary
chemical data can be found in Artinger et al. (1999),Buckau et al. (2000) and Kim et al. (1995). Comparedto the above established age ranges, large di�erences
are found. With few exceptions the calculated ages aremuch too great. Furthermore, the 14C ages vary con-siderably, re¯ecting di�erences in the models applied.
For recharge groundwaters, calculated ages range from1 to 10 ka compared to <40 a indicated by the 3Hcontent. For the transition groundwaters Gohy-182and -201, the results do not necessarily contradict the
given groundwater age ranges. Considering that 3Hcontaining recharge groundwaters are found down to24 m depth and the recharge groundwater Gohy-201 is
found at depth of 30±35 m, the model 14C agesbetween 6000 and 10 ka appear to be very high. Theenhanced DOC groundwaters are all expected to be of
Holocene origin, whereas the 14C results presented inTable 2, with few exceptions, indicate Pleistocene ori-gin. With respect to the channel brines, the 14C results
and age limits for Gohy-514 do not contradict eachother, for Gohy-193 and -653, however, this appears
to be the case.The reasons for the large discrepancies may be
sought in two sources, namely (i) the 14C concentration
in primary recharge is normally postulated to be 100pmc in contrast to more probable lower values due tosoil-CO2 from both root respiration and turnover of
older humus; and (ii) in enhanced DOC groundwaters,in addition to 14C decay and dilution of 14C with DICof inorganic sediment origin, the 14C concentration is
lowered by the mineralization of 14C free sedimentaryorganic C (SOC).
4.3. Dilution of DIC by the mineralization of SOC
In the Gorleben groundwaters three sources of DIC
need to be considered (Fig. 5). These are (i) the pri-mary recharge which is the 14C containing age indica-tive tracer originating from organic sources in rechargewater; (ii) mineralization of organic sediments; and (iii)
dissolution of sedimentary carbonate. Many 14Cgroundwater dating methods rely to a large extent onthe 13C signature to quantify the impact of dilution by14C-free DIC. Thereby, dilution is assumed to occurthrough dissolution of carbonate sediments, normallywith d13C 1 0 for marine carbonate. Upon mineraliz-
ation of SOC, DIC of organic origin with a di�erent13C signature is introduced. Consequently, the use ofd13C 1 0 to quantify the degree of dilution with 14C
Fig. 5. Major sources of DIC in groundwater where dilution
of primary recharge is governed by dissolution of carbonate
sediments and the mineralization of sedimentary organic C.13C values can vary somewhat around the numbers given. The14C value of 1100 pmc for root respiration refers to pre-
nuclear atmospheric testing conditions.
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597590
free DIC will be erroneous. Because the mineralizationof SOC appears in conjunction with in situ generation
of aquatic humic substances, the mineralization processis expected for enhanced DOC groundwaters. Ingroundwaters where precipitation or decomposition of
aquatic humic and fulvic acids has occurred, althoughthe mineralization process may have occurred theenhanced DOC concentration cannot be used as an in-
dicator.To judge the applicability of the 13C signature for
14C groundwater dating, a scheme is applied plotting
the 13C and DIC concentrations against each other (cf.below). For this purpose the 13C concentration of pri-mary recharge DIC is required. For closed conditionsisotopic fractionation need not be regarded (Taylor,
1998) and thus primary recharge DIC will have thesame 13C concentration as the source material (d13C=ÿ 27-) (Kim et al., 1995). In Fig. 6, d13C values of
shallow groundwater from the Fuhrberg and Munichaquifer systems are plotted against pH (Kim et al.,1995), together with data from recharge and transition
water of the Gorleben aquifer system (Table 1). Fuhr-berg sediments do not contain carbonates and thusDIC is only of organic origin. Furthermore no isotopic
fractionation is seen. Due to di�erent degrees of min-eral dissolution, pH varies between values of precipi-
tation and a value of approximately 6. In the Munichaquifer system 98 % of sediments consist of calcite
and dolomite. In these shallow waters (0.5 to 23 m),the primary recharge DIC leads to dissolution of equalamounts of DIC from sediments. Consequently, these
pH neutral waters have 13C concentrations exactlybetween the two sources. In Gorleben recharge andtransition waters, pH appears to become progressively
shifted to higher values both by dissolution of carbon-ate and non- carbonate containing sediments. The
Fig. 7. (a) 13C signatures plotted against the DIC concen-
trations. The dilution curve represents groundwater conditions
where mineralization of sedimentary organic C (SOC) is of
negligible magnitude. Values above the dilution curve in gen-
eral represent groundwaters considerably a�ected by the min-
eralization of SOC, accompanied by in situ generation of
dissolved organic C (DOC). (b) 13C signatures plotted against
the DIC concentrations. (a)±(d) represent di�erent possible
mixing lines (see text). Vertical shaded lines represent 0.51
mmol/L primary recharge DIC (d13C= ÿ 27-) in Gohy-181,
-182, -611 and -711, and 0.56 mmol/L DIC in Gohy-411 and -
421 (and possibly Gohy-341).
Fig. 6. 13C and pH of Gorleben recharge and transition
groundwaters and shallow groundwaters from the Fuhrberg
and Munich aquifer systems. The lower horizontal line
(d13C=ÿ 13.5-) represents equal contributions from primary
recharge DIC (d13C=ÿ 27-) and DIC from sedimentary car-
bonate (d13C=0).
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597 591
trend, however, makes it a plausible assumption that
the primary recharge is not subject to isotope fraction-ation but has a d13C value of around ÿ27-.In Fig. 7a, d13C is plotted against the DIC concen-
tration for the Gorleben groundwaters. Data fallingclose to the curve representing dilution of primary
recharge DIC with 14C free DIC from marine carbon-ate sediments may be allocated to a common geochem-ical evolution history and are not signi®cantly a�ected
by the mineralization of SOC. In contrast to this,groundwaters found above this curve are eithera�ected by the mineralization of SOC or have an ori-
gin in signi®cantly di�erent recharge conditions. InFig. 7b an alternative representation of the 13C content
versus DIC is used. The 13C signature is plottedagainst the inverse of the DIC concentrations. By thisrepresentation, mixing of two DIC sources with di�er-
ent 13C signatures fall on a straight line. The disadvan-tage is that at low DIC concentrations the scale isstretched whereas at high concentrations it is com-
pressed.In Fig. 7b, a number of possible mixing processes
are shown by straight lines. Line (a) shows that therecharge and transition groundwaters Gohy-181, -182,-611 and -711 can be represented by 0.51 mmol/l pri-
mary recharge DIC which becomes diluted by DICwith d13C=0. Enhanced DOC can be represented byaddition of DIC with a d13C value of ÿ13.5- to well
developed recharge/transition water (b). Such a sourceresults from oxidation of SOC to carbonic acid fol-
lowed by dissolution of equal amounts of sedimentarycarbonate (with d13C=0). Gohy-2227 shows a morenegative d13C value consistent with the somewhat
lower pH value than the other highly enhanced DOCsamples, i.e. the carbonic acid of organic origin is notfully compensated by dissolution of sedimentary car-
bonate. The brines cluster around line (c), representingmixing of a well developed recharge/transition water
and DIC of organic origin (d13C= ÿ 27-). The chan-nel brines, however, are separated by several km dis-tance and between recharge/transition waters and the
deep channel brines enhanced DOC groundwaters arefrequently found (Buckau et al., 1999). Therefore, theassumption of an origin of the channel brines in a
common type of well developed recharge/transitionwater, without in¯uence of mineralization of SOC
during sinking to greater depths, appears very unlikely.The shallow uplift-mixing water Gohy-341 and the
recharge waters Gohy-411 and -421 can be represented
by a common mixing line. This line (d) represents mix-ing of 0.56 mmol/l primary recharge DIC and DIC
with d13C= ÿ 13.5-. Gohy-411 and -421 originatefrom further south than the other recharge waters, anda DIC contribution from shallow mineralization of or-
ganic sediments followed by dissolution of sedimentarycarbonate may occur due to local conditions. In the
uplift-mixing water Gohy-341, located west of the saltdome, this process is much more pronounced. The
high DOC concentration of this water is also indicativeof considerable mineralization of organic sediments(Buckau et al., 1999). In comparison with the
enhanced DOC groundwaters, Gohy-341 shows amuch higher 14C concentration, showing that relativelyyoung C sources are involved in the mineralization
process (Kim et al., 1995).In summary, the mineralization of SOC has nor-
mally not been taken into account and thus di�culties
in evaluation of 14C groundwater ages are encountered.In the Gorleben aquifer system this mineralization ofSOC is of considerable importance and thus needs tobe taken into account. The 13C signature can support
evaluation of the DIC inventory, however, it cannotdistinguish between age indicative primary rechargeDIC and non-14C containing DIC of organic origin
from deep sediments. Therefore, for the enhancedDOC waters, the overall DIC concentration ratiosbetween the primary recharge and total DIC ground-
water concentrations is used. In the channel brines thegeochemical history of DIC appears unclear and thus14C groundwater dating may give questionable results.
In the up-lift mixing groundwater Gohy-341, the 14Cconcentration re¯ects generation of DIC from shallowsources and thus 14C dating appears meaningless.Another problem is related to the fact that, on using
frequently applied published methods, even young 3Hcontaining groundwaters are found to have agesbetween 1 and 10 ka. This leads to the question of the
source term for the 14C groundwater dating.
4.4. 14C source term
For groundwater dating, the 14C concentration of
primary recharge at the time of recharge is required asa starting point. The age indicative 14C concentrationof primary recharge will be governed by the relative
contributions from young sources, namely root respir-ation re¯ecting atmospheric 14C concentrations andturnover of old humus via microbial activity. Already
in the recharge groundwaters, the 14C concentration ofDIC may decrease by dilution through dissolution ofsedimentary carbonates. This process can be quanti®edvia the di�erences in the d13C values of the two
sources.The shallow waters in the recharge area may be
divided into two di�erent groups (cf. above and
Fig. 7b). This re¯ects di�erences in present localconditions. For evaluation of a source term onemust consider that (i) the absolute di�erence in the
primary recharge DIC concentrations is not verylarge, (ii) local recharge conditions may change withtime, and (iii) considerable intermixing of ground-
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597592
water takes place in this aquifer system (Buckau et
al., 1999) and thus the origin of deep groundwater
is probably a mixture from di�erent recharge lo-
cations. For evaluation of the age indicative primary
recharge DIC, 13C is used to calculate the original
recharge contribution of organic origin in all recharge
and transition waters. In Table 3, the DIC concen-
trations and d13C values are given. Applying a d13Cvalue of ÿ27- for primary recharge and zero for sedi-
mentary carbonate, admixing of DIC of inorganic ori-
gin is corrected for. With respect to the primary
recharge DIC concentration, the results for recharge
and transition groundwaters can be summarized as
0.602 0.21 mmol/L. On proceeding similarly the 14C
value for primary recharge DIC is obtained (Table 3).
A closer look on the recharge and transition
groundwaters reveals that the calculated 14C concen-
trations of primary recharge are in¯uenced by fall-out
from atmospheric nuclear testing. In Fig. 8, the 14C
concentrations of primary recharge DIC are plotted
against the groundwater 3H concentrations. In this
®gure also data from Suckow (1993) are shown,
including several samples with low 3H concentrations.
To allow comparison of data from di�erent sampling
years (1990±1994) all 3H data have been normalized
to, 1994. Linear correlation result in a 14C concen-
tration of primary recharge, corrected for atmospheric
nuclear testing. For comparison, data from Fuhrberg
groundwaters (cf. above) are also shown. The values
are scattered around the correlation for Gorleben
recharge groundwaters and thus support the ®ndings
in Gorleben.
The data points in Fig. 8 can be divided into three
groups, (1) those resulting in above linear correlation,
(2) groundwaters not considerably in¯uenced by fall-
out, and (3) some groundwaters with signi®cant 3H
concentrations deviating considerably from liner re-
lationship with 14C. The groundwaters are sampled
within a period of 4 a and thus correction for decayshould adequately re¯ect variations in recharge 3Hconcentrations. The reasons for considerable deviations
in some samples is not clear. The overall conclusion,however, is that prior to atmospheric nuclear testingthe 14C concentration of primary recharge in the twocomparable aquifer systems Gorleben and Fuhrberg is
not even close to the frequently assumed approxi-mately 100 pmc but is somewhat above 50 pmc.In summary, in Gorleben the primary recharge is
calculated to consist of 0.6020.21 mmol/L DIC with
Table 3
Gorleben recharge and transition groundwaters: DIC concentration and its 13C and 14C signatures, and 13C and 14C concentrations
for primary recharge of DIC. Primary recharge refers to DIC excluding contribution originating from sedimentary carbonate. Cor-
rection is made via 13C with a d13C of ÿ27- for DIC of organic origin (primary recharge) and dC13 of zero for DIC originating
from sedimentary carbonate
Sample DIC
(mmol/L)
d13C(DIC)
(- (rel. PDB))
14C
(DIC)
(pmc)22s
DIC
(primary recharge)
(mmol/L)
14C
(DIC (primary recharge))
(pmc)
Gohy-181 0.67 ÿ 20.6 45.720.9 0.51 59.9
Gohy-182 1.07 ÿ 13.4 27.920.6 0.53 56.2
Gohy-201 0.79 ÿ 13.6 26.522.1 0.40 52.6
Gohy-411 1.54 ÿ 17.9 54.223.3 1.02 81.8
Gohy-421 0.84 ÿ 23.1 68.524.5 0.72 80.1
Gohy-611 1.06 ÿ 13.3 26.521.7 0.52 53.8
Gohy-711 0.59 ÿ 22.4 70.421.4 0.49 84.9
0.6020.21 67.02 14.5
Fig. 8. Impact of atmospheric testing on the 14C concen-
tration of primary recharge DIC in Gorleben groundwaters.
Open symbols are not regarded for evaluation of 3H and 14C
originating from nuclear atmospheric testing. Data from the
Fuhrberg aquifer system with comparable conditions are
shown for comparison (for explanation, see text).
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597 593
a 14C content of 54.221.7 pmc, corrected for the e�ectof atmospheric nuclear testing.
4.5. Evaluation of groundwater age
As already discussed above, after recharge, the 14Cconcentration will decrease not only by decay, but alsoby dilution with DIC from dissolution of carbonate
sediment and mineralization of SOC. To visualize theimpact of overall dilution of DIC, irrespective of thedilution source, the 14C concentration of DIC is
plotted against the total DIC concentration in Fig. 9.The primary recharge is set to 54.2 pmc at a DIC con-centration of 0.60 mmol/L (see above). From this pri-
mary recharge a dilution curve is plotted for theexpected 14C concentration on the assumption that no14C decay has occurred. The results for the enhanced
DOC groundwaters fall around the dilution curve,suggesting no signi®cant 14C decay. The channel brineGohy-653 also falls on the dilution curve and thus no14C age is found. Contrary to this, Gohy-193 and -514
fall somewhat below the dilution curve. The deviationof the experimental value from the dilution curveresults in the 14C age of approximately 4 ka for Gohy-
514 and approximately 10 ka for Gohy-193.The uplift-mixing Gohy-341 falls well above the di-
If applying the DIC contribution from organic origin(cf. primary recharge), 3.07 mmol/L DIC with a 14Cconcentration of 44.7 pmc is found. The high DICconcentration is the result of mineralization of organic
sediment, accompanied by a high DOC content. This
14C concentration is rather close to the 54.2231.7 pmcfound for the primary recharge of Gorleben ground-
waters. The NaCl content is similar to the enhancedDOC groundwaters Gohy-2211 and -2227 (from 80±85and 128±130 m depth, respectively and the 3H concen-
tration is similar to the enhanced DOC groundwatersGohy-572, -612 and -2211 (Table 1). Therefore, onemay assume that at least major parts of this ground-
water is relatively old. The DIC, however, is obviouslygenerated at shallow depth, and thus 14C groundwaterdating of this water is not applicable.
In Table 4, the groundwater age ranges and resultsfrom published methods (cf. Table 2) are showntogether with the present results. Also shown areresults for the enhanced DOC groundwaters applying
a conservative value of 83.1 pmc for the 14C content ofrecharge DIC (Buckau et al., 1997) and channel brinesages resulting from radiocarbon dating on aquatic ful-
vic acids (Artinger et al., 1996). In general the resultsfrom published 14C methods show unrealistically highgroundwater ages, frequently contradicting stable iso-
tope data and 3H. The present analysis shows that thetransition and enhanced DOC groundwaters at depthsbetween 30 and 137 m are so young that decay of 14C
cannot be quanti®ed accurately.Present 14C dating results for the channel brines are
not consistent with the given age limits obtained by 2Hand 18O. Application of 14C dating on aquatic fulvic
acids (Artinger et al., 1996) gives somewhat lowgroundwater ages compared to 2H and 18O data, butthe trend seems to be reasonable (Table 4). The dis-
agreement between the 14C dating on DIC and onaquatic fulvic acid leads to the conclusion that a re-liable 14C groundwater dating for the channel brines
presently is not possible. The isotope signatures in theGorleben channel show various layers of groundwaterof Holocene and Pleistocene origin and layers of var-ious salinity are assumed to ¯ow in di�erent directions.
The contradicting results of groundwater dating re¯ectthe unclear geohydrological and geochemical situationin the Gorleben channel.
5. Summary and conclusions
The developed 14C groundwater dating method
makes use of (i) a site speci®c pre-atmospheric nucleartesting primary recharge DIC as the age indicative tra-cer, and (ii) overall dilution of primary recharge DIC
in deep groundwater. The DIC concentration of thesite speci®c primary recharge and its 14C concentrationis calculated via the d13C values of recharge and tran-
sition groundwaters, that are not signi®cantly in¯u-enced by the mineralization of SOC. The impact offall-out 14C in primary recharge is corrected for via
Fig. 9. 14C concentrations of DIC in Gorleben enhanced
DOC, uplift-mixing and channel brine groundwaters plotted
against the total DIC concentration. The dilution curve rep-
resents the 14C concentrations expected for no 14C decay.
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597594
3H. The site speci®c DIC recharge, relevant for pre-at-
mospheric nuclear testing conditions, is found to be
0.602 0.21 mmol/L DIC with a 14C concentration of
54.221.7 pmc.
Application of the present approach to the
enhanced DOC groundwaters sampled between 35
and 137 m depth leads to the conclusion that these
groundwaters are so young that signi®cant 14C
decay cannot be detected. These results are very di�er-
ent from those found by application of published
methods (groundwater ages reaching up to 31 ka). For
these enhanced DOC groundwaters, the present study
leads to groundwater ages in compliance with the age
limits derived by the 2H and 18O signatures.
In the Gorleben channel, groundwater dating of
brines by di�erent methods (including the present
study) gives widely di�erent and inconsistent results.
In view of the previously known discrepancies and irre-
gularities, including in isotope data and lack of suc-
cessful modeling of hydrological conditions in the
Gorleben channel, the large di�erences in groundwater
dating results are not surprising. The present studyunderlines that a reliable groundwater dating can only
be achieved when the hydrological conditions for thestudied site are better known and adequate knowledgeof geochemical processes is available. In the shallow
peat-bog like uplift-mixing groundwater Gohy-341, the14C concentration re¯ects generation of DIC from nearsurface organic sources but not the origin of thegroundwater. Consequently groundwater dating by 14C
is not applicable.
Acknowledgements
The authors acknowledge the ®nancial support of
the `Bundesministerium fuÈ r Bildung, Wissenschaft,Forschung und Technologie`. We also acknowledge the`Bundesamt fuÈ r Strahlenschutz` for making the Gorle-
ben data base available. Thanks are due to H. Siela�and D. Wesselow of the DBE at the Gorleben site fortheir assistance in the groundwater sampling, and C.
Table 4
Age of investigated Gorleben groundwaters: ranges of groundwater age from 3H, 2H and 18O (Table 2) and results from appli-
cation of published, frequently used 14C groundwater dating methods (Table 2), results from application of primary recharge of
0.60 mmol/L DIC with a 14C concentration of 83.1 pmc (not accounting for fall-out) (Buckau et al., 1997), 14C dating of aquatic
fulvic acid (Artinger et al., 1996) and results of the present study (groundwater age in a)
Sample Sample No. Age range
(cf. Table 2)
Frequently applied methods
(cf. Table 2)
Buckau et al.,
1997
Artinger et al., 1997 This work, 1996
Recharge area
1 Gohy-181 < 40 4000±8000 recharge
2 Gohy-411 < 40 1000±6000 recharge
3 Gohy-421 < 40 2000±6000 recharge
4 Gohy-611 1 40 5000±10,000 recharge
5 Gohy-711 < 40 2000±5000 recharge
Transition area
6 Gohy-182 40±10,000 5000±9000 nsda
7 Gohy-201 40±10,000 6000±10,000 nsd
Enhanced DOC
8 Gohy-412 40±10,000 8000±19,000 3000 nsd
9 Gohy-492 40±10,000 25,000±31,000 5000 nsd
10 Gohy-572 40±10,000 4000±24,000 6000 nsd
11 Gohy-573 40±10,000 18,000±26,000 3000 nsd
12 Gohy-612 40±10,000 27,000±31,000 5000 nsd
13 Gohy-2211 40±10,000 17,000±23,000 1 0 nsd
14 Gohy-2227 40±10,000 19,000±24,000 1000 nsd
Channel brines
15 Gohy-193 1 10,000 28,000±30,000 4000 10,000
16 Gohy-514 < 12,000 19,000±23,000 11,000 4000
17 Gohy-653 1 10,000 16,000±20,000 6000 nsd
Uplift mixing
18 Gohy-341 mixb 7000±10,000 not applicable
a nsd: no signi®cant decay of 14C detected.b mix: mixture of deep and shallow groundwater where DIC is generated mainly at shallow depth.
G. Buckau et al. / Applied Geochemistry 15 (2000) 583±597 595
Kardinal (Institute of Radiochemistry, TU MuÈ nchen)and H. Halder (GSF-Institute of Hydrology) for their
technical assistance. The isotope measurements werecarried out under the supervision of W. Rauert and P.Trimborn at the GSF-Institute of Hydrology. The
Authors would also like to thank F.J. Pearson Jr.(PSI, Villingen) for his valuable comments.
References
Appelo, C.A.J., Postma, D., 1996. Geochemistry, ground-
water and pollution. A.A. Alkema, Rotterdam.
Aravena, R., Wassenaar, L.I., Plummer, L.N., 1995.
Estimating 14C groundwater ages in a methanic aquifer.
Water Resour. Res. 31 (9), 2307±2317.
Artinger, R., Buckau, G., Kim, J.I., Geyer, S., Wolf, M.,
1996. In¯uence of sedimentary organic matter on dissolved
fulvic acids in groundwater, signi®cance for groundwater
dating with 14C in dissolved organic matter. In:
Proceedings of the IAEA Symposium Isotopes in Water
Resources Management, 20±24 March 1995. IAEA,
Vienna, pp. 52±72.
Artinger, R., Buckau, G., Kim, J.I., Geyer, S., Fritz, P.,
Wolf, M., 1999. Characterization of groundwater humic
substances: In¯uence of sedimentary organic C. Appl.