Water Research 36 (2002) 3989–4000
Sulphur isotopes as tracers of the influence of potash mining in groundwater salinisation in the Llobregat Basin (NE Spain) N. Otero*, A. Soler " Departament de Cristal.lografia, Mineralogia i Diposits Minerals, Facultat de Geologia, Universitat de Barcelona, Mart!ı i Franqu"es, s/n, 08028, Barcelona, Spain Received 9 August 2001; accepted 19 February 2002
Abstract Conventional chemical data for spring and river waters are presented together with sulphur isotopic data for dissolved sulphate to elucidate the source of water salinisation in the middle section of the Llobregat River. As dilution processes do not affect sulphur isotopic composition, the analysis of d34S of dissolved sulphate in waters provides an excellent tool for quantifying the environmental impact caused by the mining activity existing in the area. The d34S of dissolved sulphate from mining effluents and saline springs unrelated to mining activity was analysed. The results obtained range from +18% to +20% (VCDT) for mining effluents and from +10% to +14% (VCDT) for natural saline springs. These values are in accordance with the pattern of sulphur isotopic composition of sulphates from the evaporite materials of this area. This distinctive isotopic composition has allowed us to determine the origin of salinity in those cases in which chemical features are not conclusive. In addition, two fertilisers widely used in the studied area are chemically and isotopically characterised and their contribution to groundwater salinisation is assessed. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Sulphur isotopes; Potash mining; Water salinisation; Fertilisers; Contamination
1. Introduction In the middle section of the Llobregat Basin (Northeastern Spain) several potash mining works exist in the ! localities of Sallent, Suria, Balsareny and Cardona (Fig. 1). This activity produces large salt mining tailings, which are stored around the mining sites, with no waterproofing. Some aquifers near the mining zones are highly salinised. The origin of the salinity is controversial, as it can be related to natural water interaction with saline formations, or it could be caused by contamination from mine tailing effluents. Fertilisers could also contribute to groundwater salinisation, as agriculture is an important economic activity in the Llobregat Basin.
*Corresponding author. Tel.: +34-93-402-1345; fax: +3493-402-1340. E-mail address:
[email protected] (N. Otero).
Sulphur isotopes of dissolved sulphate can be used to identify natural sources of sulphate in surface waters [1– 6] and in groundwater [7–9]. In some cases they can be applied to distinguish between natural and anthropogenic sources [10–16] and to trace processes in the soil [17,18]. In a previous paper [19], we reported differences between sulphur isotopic composition of natural and anthropogenic sources in the Llobregat River waters large enough to consider d34S as a good tool to discriminate their different origins. This tool, however, is more useful when the studied area is small since the inputs can be better constrained and more limited. On the basis of our previous results, the aim of the present paper is to determine the origin of groundwater salinisation in a specific area of Llobregat Basin with some saline springs, using the sulphur isotope composition of dissolved sulphate as well as water chemistry.
0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 3 - 1 3 5 4 ( 0 2 ) 0 0 1 2 5 - 2
3990
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
Fig. 1. Llobregat River Basin, showing major tributaries and sample location.
2. Study area The area studied is the middle section of the Llobregat River. For more information on the river’s characteristics see [19]. In this area, besides the main flow, there are three tributaries, all of which have several evaporitic outcrops in their catchments (Fig. 2). The Cardona Diapir, which appears in the Cardener River basin, is composed dominantly of halite, sylvite, carnallite and
gypsum of Late Eocene age. In the studied basin different sulphate formations outcrop: gypsum of Oligocene age [20], and both anhydrite and gypsum of Late Eocene age [21]. These evaporite materials cause high natural values of salinity (1–4%wt) in some of the small tributaries of the rivers. On the other hand, the salinity of the Llobregat River has increased significantly since the early 1920s [22] due to the extensive development of potash and salt mining.
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
3991
Fig. 2. Simplified geological map of the Llobregat Basin, see Fig. 1 for sample location.
The potash mine tailings mainly consist of halite with minor amounts of sulphates (gypsum, polihalite and anhydrite) and are considered by law as a resource and not as waste, as halite can be reused for the chlorine industry. Because of this, there are large mine tailings, some without waterproofing (Plate 1), stored in the surroundings of the mining sites (Balsareny, Cardona, ! Sallent and Suria). The mining industry also dumps scrap metals into the tailings, which cover an area of
98 ha and the estimated mass is 61 Mt, with an annual increase of about 2.7 Mt [23]. Since 1989, a brine collector drives the brines from the potash mining zones, and from several springs, draining evaporitic materials, to the sea, discharging it 10 km before reaching the coast. However, not all saline springs are collected and some flow into the Llobregat River. Furthermore, two of the main fertilisers used in the agricultural activity of the region are ammonium
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
3992
Plate 1. Mine tailings located at site 4.
sulphate and the NPK 5.7.10 (5% of N, 7% of P2O5 and 10% of K2O) that supply nitrate, sulphate, potassium, etc. to groundwater.
3. Methodology In order to characterise the water chemistry and the sulphur isotopic signature of dissolved sulphate, saline samples (conductivity up to 15 mS/cm) from springs and river waters were collected. Five water samples correspond to springs with a salinity of natural origin, 10 are from waters flowing through potash mine tailings, and 10 from springs where the origin of salinity is uncertain. The samples were collected between July 1997 and November 2000. Sampling localities are shown in Fig. 1. Samples located at sites 1, 2, 6, 7 and 10 are from springs that have highly variable discharges, whereas samples located at site 4, 5, 8 and 9 are from springs that have relatively constant discharge. Samples from site 3 are from a Llobregat tributary with irregular flow regimen. The samples from mine tailing effluents sampled at sites 2, 3 and 4 are drained to the salt collector; however, at the time of sampling site 2, the effluent flowed into the Cardener River. The rest of the samples from mining effluents (site 1), natural origin (sites 5 and 6) and uncertain origin (sites 7, 8, 9 and 10) are not drained to the salt collector. Physicochemical parameters (pH, temperature and dissolved O2) were measured in situ. Rainwater samples were collected in a pluviometer between October 1997 and December 1998 (see location in Fig. 1). In order to avoid sulphate reduction during sample collection, the dissolved sulphate was precipitated as BaSO4 by adding a solution with excess of BaCl2 2H2O in the pluviometer. To constrain the importance of the possible supply of fertilisers to the global isotope composition of
dissolved sulphate, sulphur isotope composition of two of the main fertilisers used in the area was also analysed. Before analysis, samples were filtered with a Millipores filter of 0.45 mm pore size (fertilisers were previously dissolved in distilled water). Major anions (NO3 , SO4 2 and Cl ) were analysed by HP liquid chromatography. Concentrations of the major cations (Ca, Mg, Na, Sr and K) were determined by ICP-OES, and the trace elements (Mn, Ba, Zn, Cu, Cr, As, Se, Cd, V, Hg, Pb and Co) were analysed by ICP-MS; due to the high chloride content, most samples needed a dilution and therefore, most trace elements were under the detection limits. Alkalinity was measured by titration. For sulphur isotope analysis the dissolved sulphate was precipitated as BaSO4 by the addition of BaCl2 2H2O. The sulphur isotopic composition was determined with an elemental analyser (Carlo Erba 1108) coupled with a mass spectrometer (Finningan Matt delta C). Notation is expressed in terms of d34S per mil relative to the Vienna Canyon Diablo Troilite (VCDT) standard. The isotope ratios were calculated using the NBS-127, IAEA-S1, IAEA-S2, IAEA-S3 and internal laboratory standards. Reproducibility of the samples calculated from standards systematically interspersed in the analytical batches is 70.2%. All chemical and sulphur isotope analyses were carried out at the Serveis Cient!ıfic T"ecnics (Universitat de Barcelona).
4. Results and discussion 4.1. Major and trace elements Results of pH, conductivity, major anions and sulphur isotopic composition of dissolved sulphate are shown in Table 1; the results of major cations, trace elements and estimated flow are shown in Table 2. All spring and river water samples are considered as ‘‘Cl– Na’’ waters (Fig. 3). According to their relative cation abundance, two groups of samples may be distinguished: one with Na>Ca>K, and another with Na>K>Ca. The Na/K ratio of the first group ranges from 100 to 500 whereas in the second it is one order of magnitude lower (2–20) (Table 2). The Na versus K and Na versus Mg diagrams (Figs. 4a and b) show three clusters of samples. The first, A, has Na contents from 0.1% to 1.5%, K ranging from 10 to 500 ppm and Mg from 200 to 1500 ppm. The second, B, is characterised by B11% of Na, from 0.5% to 1.5% of K and B0.1% of Mg. The third, C, has B5% of Na and high contents of K (2–3%) and Mg (B0.6%). All water samples having a salinity of natural origin plot in group A. Four samples (8a, 9a, 9b and 10a) with salinity from uncertain sources, also plot within group A, suggesting a natural origin as well. The rest of the samples with salinity from uncertain sources
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
3993
Table 1 Anion concentration and sulphur isotopic composition of dissolved sulphate Site Sample Location
Date pH (month-yr)
Cond Cl (mS) (201C) (mg/l)
SO4 NO3 (mg/l) (mg/l)
1 2 3
Apr-99 Mar-98 Mar-98 Jul-97 Apr-98 May-98 Aug-98 Aug-98 Nov-00 Nov-00 Mar-98 Mar-98 Apr-99 Apr-99 Apr-99 Mar-98 Apr-99 Apr-99 Apr-99 Jun-98 Dec-99 Dec-99 Mar-98 Jun-98 Apr-99 Dec-98 — —
562,000 584,000 604,000 n.d. 188,400 173,700 158,700 158,000 n.d. n.d. 15,050 66,100 75,600 17,200 86,800 388,000 535,000 491,000 456,000 252,000 318,000 40,100 18,190 14,040 39,600 n.d. n.d. n.d.
154 10,340 125 90 9950 194 74 9980 215 o200 9480 n.d. o100 12,700 153 o100 11,700 217 o200 8030 204 o200 7300 213 o100 5310 399 o100 10,800 147 2 2190 298 34 3420 352 22 3780 200 11 1000 330 15 4970 140 42 5370 377 220 6840 170 131 6900 421 o200 4100 316 o100 3560 383 o100 5140 365 o20 1420 278 17 300 210 5 288 212 7 481 212 1 2 n.d. o100 402,800 n.d. 4600 796,500 n.d.
4 5 6
7
8 9 10 P
1a 2a 3a 3b 3c 3d 3e 3f 4a 4b 5a 6a 6b 6c 6d 7a 7b 7c 7d 7e 7f 8a 9a 9b 10a P FK FA
Mining lixiviates (Vilafruns) Mining lixiviates (Cardona) Rierol Salat Rierol Salat Rierol Salat Rierol Salat Rierol Salat Rierol Salat Mining lixiviates Mont Salat Mining lixiviates La Botjosa Gorg Salat, surge Olo! Stream, surge 1 Olo! Stream, surge 1 Olo! Stream, surge 2 Olo! Stream, surge 3 Sallent, surge 1 Sallent, surge 2 Sallent, surge 3 Sallent, surge 1 Sallent, surge 1 Sallent, surge 1 ! surge Callus, Hortons Stream, surge Hortons Stream, surge ! Suria, surge C-1410 Pluviometer Fertiliser NPK (5.7.10) Ammonium sulphate
7.4 7.4 7.4 n.d. 7.4 7.4 7.3 7.4 n.d. n.d. 7.3 7.4 7.7 7.7 7.6 6.8 7.6 7.0 7.2 7.2 7.2 7.2 7.7 7.8 7.8 n.d. n.d. n.d.
162,000 176,300 177,800 139,200 160,100 136,800 189,300 175,300 213,000 189,200 3020 19,700 22,300 4080 23,500 109,200 137,300 141,300 130,000 75,300 92,900 13,500 4850 5560 11,300 o2 51,700 30,500
HCO3 CO3 (mg/l) mg/l o1 o1 o1 n.d. 17.9 o1 o1 o1 o1 o1 o1 o1 o1 o1 o1 o1 43.2 o1 o1 o1 o1 o1 o1 o1 o1 n.d. n.d. n.d.
d34S (%) VCDT 20.1 17.9 18.9 19.4 18.5 19.3 19.4 19.6 20.0 18.3 10.2 13.8 12.8 11.8 13.5 18.5 19.8 18.2 18.9 18.2 19.1 8.6 13.7 13.7 2.4 7.2 6.9 1.1
n.d.=not determined. Fertiliser concentrations are in mg/kg.
form group C. Group B includes water samples from the Salat Stream (draining the Cardona diapir as well as mine tailings, site 3) and samples from mine tailing effluents (1a, 2a and 4b). Sample 4a from mine effluents plots outside these groups. The mine effluents collected at site 4a drain active mine tailings; the effluents collected at sites 1, 2, 3 and 4b are from abandoned mine tailings. As solubility of silvite and carnallite is greater than solubility of halite, the active and the abandoned mine tailings would have different K and Mg contents. On a Na/Mg versus Na/K diagram (Fig. 4c) samples plot in two distinct trends: one with a wide Na/K range and low Na/Mg ratio which are the natural springs; and a second trend with a low Na/K ratio and high Na/Mg variability which includes samples from mining effluents. The samples with salinity of uncertain origin plot in the lower Na/K and Na/Mg ratio region, and cannot clearly be associated either to a natural or to a mining origin, although samples 9a and 9b seem to fit in the trend defined by the salinity of the natural springs.
K and Mg contents allow the distinction between samples of natural origin and samples of mining effluents because the potash unit is mainly made up of sylvite (KCl) and carnallite (KMgCl3 6H2O). Nevertheless, K and Mg data are not conclusive indicators of the origin of salinisation in uncertain cases since (a) there are other sources for K and Mg in the basin such as NPK fertilisers (nitrate–phosphate–potassium) and/ or industry pollutants; and (b) Na/K and Na/Mg ratios from the mine tailings are time dependant as solubility of silvite and carnallite is greater than solubility of halite. Therefore, as suggested above, K and Mg contents in active mine tailings are greater than in abandoned mine tailings. In the chloride versus sulphate diagram (Fig. 5a) the natural and mining samples are well differentiated: natural samples have o2% of Cl and o0.5% of SO4 2(group A), whereas mining samples contain B15% of Cl and variable amounts of SO4 2 (group B). As previously suggested by Na–K–Mg contents (Fig. 4a and b), the Cl /SO4 2ratios in the samples 8a, 9a, 9b and 10a also point to a natural origin (group A). The rest of
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
3994
Table 2 Major and trace cation contents and estimated flow Site
Sample
Na (mg/l)
K (mg/l)
Mg (mg/l)
Ca (mg/l)
Sr (mg/l)
Mn (mg/l)
Zn (mg/l)
Cu (mg/l)
Pb (mg/l)
Na/K
Estimated flow
1 2 3
1a 2a 3a 3b 3c 3d 3e 3f 4a 4b 5a 6a 6b 6c 6d 7a 7b 7c 7d 7e 7f 8a 9a 9b 10a P FK FA
108,200 111,900 118,000 112,900 113,550 110,450 116,750 111,950 85,260 133,400 1780 13,170 13,810 2880 15,630 47,970 66,520 69,030 57,810 42,870 45,060 4640 2580 2680 4500 2 3834 388
4850 12,470 11,590 7730 11,090 11,480 14,510 13,530 54,050 4310 14 54 62 12 30 27,320 26,530 25,350 29,830 21,280 22,470 548 24 22 310 o0.5 66,380 o250
725 1431 886 439 794 852 691 669 18,540 974 204 287 237 108 298 6570 5516 6161 7444 5572 5154 1408 210 225 896 0.4 7180 o10
1242 504 789 1012 754 683 955 956 1177 749 886 932 1001 360 1406 1095 1923 1366 1198 1013 1215 2089 535 551 1452 3.1 120,090 266
30 20 19 31 19 16 16 16 9 31 17 30 29 11 30 23 36 27 23 21 24 71 15 15 61 0.01 201 2
0.03 3.33 0.86 0.87 0.72 0.92 0.79 0.75 12.64 0.31 0.02 0.14 0.10 o0.05 o0.05 7.92 2.49 9.48 5.50 6.12 7.24 0.04 o0.001 o0.001 o0.005 0.004 n.d. n.d.
o0.01 o0.01 9.49 0.10 0.19 o0.01 o0.01 o0.01 0.81 o0.5 0.13 1.35 o0.01 o0.05 o0.01 1.93 o1 2.27 o1 o0.5 3.15 0.64 0.02 o0.05 o0.05 0.03 n.d. n.d.
o0.01 2.70 o0.01 0.14 0.19 o0.01 o0.01 o0.01 o0.5 o0.5 0.01 o0.2 o0.1 o0.05 o0.1 0.26 o1 o1 o1 o1 o1 o0.1 o0.01 o0.01 o0.05 o0.01 n.d. n.d.
0.09 o0.01 0.27 o0.01 o0.01 0.20 0.14 0.14 0.1 o0.05 o0.01 0.02 o0.01 o0.05 o0.01 0.89 1.51 3.69 1.08 0.41 4.98 0.01 o0.001 o0.001 o0.005 o0.001 n.d. n.d.
22 9 10 15 10 10 8 8 2 31 128 243 223 237 517 2 3 3 2 2 2 8 107 123 15 — 0.06 —
0.003 l/s 0.5 l/s 0.034 m3/s 0.034 m3/s 0.034 m3/s 0.034 m3/s 0.034 m3/s 0.034 m3/s 3 l/s 0.005 l/s 8 l/s 0.016 l/s 0.003 l/s 0.003 l/s 0.016 l/s 0.016 l/s 0.003 l/s 0.004 l/s 0.003 l/s 0.002 l/s 0.016 l/s 0.25 l/s 0.008 l/s 0.006 l/s 0.002 l/s — — —
4 5 6
7
7 8 9 10
n.d.=not determined. Fe, Ni, Co, B, Cr, As, Se, Cd, V and Hg were under the detection limit (between 1 ppb and 5 ppm depending on the dilution needed). Fertiliser concentrations are in mg/kg.
Fig. 3. Piper plot of major ions for samples of natural (K), mining (m) and uncertain (’) origin.
the samples with salinities of uncertain origin (7a–7f, group C) have intermediate values (B10% of Cl and B0.5% of SO4 2) suggesting a mixing between samples from natural springs (A) and mining sources (B).
No clear correlation between nitrate and sulphate contents is observed (Fig. 5b), which suggests that nitrate comes from a source unrelated to the mine tailing effluents. As nitrates are not present in the bedrock, they must be introduced in the waters by some human activity and therefore they might be tracers of anthropogenic pollution. In the studied area the main source of nitrates are fertilisers. The nitrate content in the analysed water samples is highly variable and depends largely on the sampling period (sites 3, 6, 7 and 9 were sampled several times). This variability could be explained by a non-constant supply of nitrate in the waters, although denitrification processes in subsurface waters cannot be ruled out. Several water samples are characterised by high contents of heavy metals (Fig. 5c), especially Mn, Zn, Pb and Cu (up to 12 ppm of Mn, 9 ppm of Zn, 5 ppm of Pb and 3 ppm of Cu). Although there is no correlation between chloride and heavy metal concentrations, all of the samples with metal contents up to 4 ppm have chloride contents above 7.5% (Fig. 5c). On the other
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000 60000
3995
20000 4a
4a
40000
C 30000
7a
10000 0
25000
10000
C
B
7e 3f 3e 2a 3a 3d 3c 1a 3b
9b 9a 6c 10a 6a 5a 8a 6d6b A
0
(a)
7d 7b 7c
7f
20000
15000
Mg (ppm)
K (ppm)
50000
5000 4b
0 50000
75000
100000
9b A 9a 6c 8a 6d 5a 10a 6a 6b
0
125000
25000
7a 7e 7f
7d
B 2a
3 a to f
1a
50000
(b)
Na (ppm)
7c 7b
75000
100000
4b
125000
Na (ppm)
300 3b
Na/Mg
200
100
3e 3f 3a 1a 4b 3c 3d 2a
4a 7 a
to
f
6b 6a 6c
9a 9b 5a 8a 10a
0 0
100
(c)
200
6d
300
400
500
600
Na/K
Fig. 4. (a) Plot of Na versus K, (b) plot of Na versus Mg, (c) plot of Na/K versus Na/Mg for (K) samples of natural origin, (m) samples of mining origin and (’) samples of uncertain origin.
hand, mine tailing effluent P samples show variable amounts of heavy metals ( [Mn, Zn, Pb, and Cu] from 0.1 to 13 ppm), the origin of which could be related to the local presence of scrap metal observed in the mine tailings, which could explain also the heterogeneity of these metal contents. Consequently, the geochemical signature of the potash tailing effluents is not only the chloride and potassium contents but in some cases, the presence of heavy metals as well. The samples from uncertain origin collected at site 7 also show high values P in heavy metals ( [Mn, Zn, Pb, and Cu] from 4 to 15 ppm) suggesting a mining origin for these samples. However, using heavy metals as tracers of mining effluent pollution is not a conclusive tool, as there are several sources that can contribute heavy metals to groundwater, e.g. fertilisers and sewage sludge usually have high contents of Zn, Cu, and Pb ([24], among others). Using conventional chemistry, samples with salt content of natural or mining origin can be well differentiated. In some cases, where salinity has an uncertain origin (samples 9a and 9b) they seem to fit with the natural salinity values but in others (samples 8a
and 10a), despite their chemical similarity to the natural values, they exhibit an unusually low Na/K ratio. The remaining water samples with salinities of uncertain origin (collected at site 7) are not clearly related to a natural or a mining source; however, the high contents in chloride, potassium and heavy metals in these samples suggest an influence of mine effluents. In order to confirm these results and to clarify the origin of salinity in the studied area, a systematic isotope analyses of the studied samples was carried out.
4.2. Sulphur isotopes 4.2.1. Sources and isotopic composition of sulphate The natural sources of dissolved sulphate in the Llobregat River are evaporites from the bedrock and rainwater. The sulphur isotopic composition determined in sulphate from rainwater has a mean value of +7%, (Table 1). However, the input of sulphate from rainwater is negligible as the mean annual precipitation in the studied area is between 500 and 750 mm and the sulphate concentration measured is below 2 ppm.
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
3996 250000
250 7b
4a 3e
Cl (ppm)
3f
150000
C 7d 100000
7c 7b
B 3a 4b 2a 1a 3b
200 3c
NO3 (ppm)
200000
3d
7a 7f
7e
50000
9b 9a
0 0
2000
6a 6b 6d
7c
100
2a 3a
50
A
9a 9b 6c 5a 10a 8a
0 4000
6000
8000
10000 12000 14000
0
(b)
SO4 (ppm)
Σ [Metals] (Mn+Zn+Pb+Cu) ppm
(a)
10a 8a 6c 5a
1a
150
16
2000
7a 6a 6b 6d 7e7d 7f 4a
4000
6000
3f 3e
8000
3b
4b
3d
3c
10000 12000 14000
SO4 (ppm)
7c
7f
14
4a
12 7a
3a
10 8
4 2 0
7d
7e
6 9b 9a 6a 6d 8a 5a 10a
0
2a
7b 3d
6b 6c
50000
100000
(c)
3b
3c 3f 1a
150000
3e 4b
200000
250000
Cl (ppm)
Fig. 5. (a) Plot of sulphate versus chloride concentration, (b) plot of sulphate versus nitrate concentration and (c) plot of chloride concentration versus sum of metals (Zn, Pb, Cu and Mn) for samples of natural (K), mining (m) and uncertain (’) origin.
The different outcropping evaporites have distinctive ! sulphur isotopic composition. According to Cendon [25], the isotopic composition of sulphate from the evaporite sequence of the Upper Eocene in the studied area shows two main stages during the development of the basin: marine and continental. The first stage corresponds to the formation of the Marginal Sulphates, the Basal Anhydrite and Lower Halite Units. The sulphur isotopic composition of the Marginal Sulphates is closely constrained around +21.670.4%. The Basal Anhydrite and Lower Halite Units are characterised by a wider range of d34S values—between +20% and +23%—which overlap the Basal Anhydrite value. The second stage corresponds to the formation of the Potash and the Upper Halite Units. The d34S of the Potash Unit ranges from 21.3% to 18.2% and decreases to +14% during the precipitation of the Upper Halite Unit. On the other hand, the sulphur isotopic composition of Oligocene sulphates from the Anoia Lacustrine Basin is more d34S depleted, ranging from +11.4% to +12.2%, due to the recycling of Triassic evaporites [26]. Consequently, the expected d34S natural values of dissolved sulphate in the studied waters would be related to the sulphur isotope composition of the drained
evaporite minerals. In places where the Potash Unit outcrops, such as in the Cardona diapir, the d34S of dissolved sulphate would range from +18% to +21%, in accordance with the d34S of gypsum from the Potash Unit. In all the others sites, these natural values would be in accordance with the values of the Upper Halite Unit (around +14%), except for site 5, where the natural value is around +12% (Anoia Lacustrine Basin). The main anthropogenic sources of sulphate in the studied waters are mining effluents and fertilisers. The d34S of the dissolved sulphate from mine tailing effluents (samples 1a, 2a, 4a and 4b) ranges from +18% to +20% in accordance with the isotope values of the potash ore. The sulphur isotopic composition of the fertilisers analysed is 1.1% for ammonium sulphate, and +7% for NPK 5.7.10. 4.2.2. Isotopic composition of dissolved sulphate Sulphur isotopic composition of sulphate dissolved in waters flowing through different rock units is shown in Table 1. Sulphate in natural springs is characterised by d34S values from +10% to +14%, consistent with the isotope composition of the evaporite outcropping
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
3997
24 22 20
4a
7d
7f 7a
7e
18
1a 7b Lower Halite 3f 3b 3d and Potash 3a 3c Units 3e 4b 2a 7c
16
12
δ 34S ( 0/00)
6a
9a 9b
14
6b
6d
Upper Halite Unit
6c Oligocene Gypsum
5a
10 8a
8
NP K
6 4 10a
2 0
materials. Thus, samples 6a–6d (site 6) have d34S values around +13% in agreement with evaporites from the Upper Halite Unit. Sample 5a (site 5), from waters flowing through evaporites of the Anoia Lacustrine System, has a d34S=+10.2% close to the isotope composition of the gypsum formation of Oligocene age (d34S=+11%). Water samples collected at site 3 (3a–3f) flow through mine tailings and evaporites from the Cardona diapir (Lower Halite and Potash Units) and their d34 SSO4 ranges from +18.5% to +19.6% (Plate 2). As sulphur isotopic composition of bedrock and mine tailings in this site is similar (d34S=+1970.5%), the sulphate contribution to waters of mining tailings is, in this case, impossible to evaluate. The variation in d34S versus [SO4] of the analysed samples is presented in Fig. 6a. Samples with the highest sulphate concentration have the highest d34S values but low isotopic variability, whereas samples with lower sulphate concentration have a broader range of isotopic compositions. Data delineate a distribution compatible with a mixing process. Different mixing curves between end-members, natural waters, mining effluents and fertilisers have been calculated and plotted with the water results (Fig. 6b). Assumed isotopic values and sulphate concentrations for mining tailing effluents (M) are d34S=+19.5% and SO4=12000 ppm; and for natural springs (N) d34S=+13.5% and SO4=5000 ppm. The concentration of the fertiliser end-members is estimated following the values obtained in small streams of the Llobregat River controlled by fertilisers [19]. The maximum concentration of dissolved sulphate in these streams is 500 ppm, so the assumed values for ammonium sulphate fertiliser (FA) are d34S= 1.1% and SO4=500 ppm; and for NPK fertiliser (FK) d34S=+7% and SO4=500 ppm.
0
(a)
5000
10000
15000
SO 4 (ppm )
20
30
60 60 90
30 60
16
δ34S (0/00)
Plate 2. Headwaters of the Salat River, draining the Cardona Diapir as well as mine tailings.
Am. Sulph.
-2
90 30 60 30
12 90
90
Natural Mining Uncertain Model FK-M Model FA-M Model N-M Model FK-N Model FA-N Am. Sulph . NPK
60
8 90
4
0
(b)
0
4000
8000
12000
SO 4 (ppm )
Fig. 6. (a) Plot of d34S versus SO4 concentration for samples of natural (K), mining (m) and uncertain (’) origin. Dashed lines show the d34S values for the main inputs. (b) Mixing models for d34S versus SO4 concentration (w/w). The end members considered are FA (fertiliser: ammonium sulphate), FK (fertilisers: NPK), M (mining effluents), N (natural surges). The calculated models are FA–M, FA–N, FK–M, FK–N and N–M.
Water samples of uncertain origin collected at site 7 (7a–7f) are from springs that used to flow into the Llobregat River, but due to a flow increase, from 0.003 l/s (March 1998) to 0.1 l/s (January 2000), have recently been taken up by the salt collector to avoid river water salinisation. These springs are located in a Quaternary fluvial terrace overlying Tertiary materials, where significant mining tailings are deposited over the
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
Quaternary sediments. The d34S of dissolved sulphate for these springs has a mean value of +18.8%. This isotopic composition is in accordance with the values of the potash unit implying that the waters are flowing through the mine tailings. The d34S of these samples can be achieved by mixing 35% sulphate from mine effluents and 65% sulphate of natural origin (Fig. 6b, model N– M). If the model considered is a mixing between mine effluents and fertilisers (models FA–M or FK–M), the proportion of sulphate from mine effluents is 25%, and 75% fertilisers. These results are especially important since, unlike conventional chemistry, the sulphur isotopic signature of the waters is not sensitive to high dilution of the original brines. As this result indicates, even when an anthropogenic source is highly diluted, and cannot be detected by classical chemistry, its sulphur isotopic signature may be used as a fingerprint. Sample 8a has a sulphur isotope value of +8%. Although the Na/K ratio points to salinity related to the mining activity (Fig. 4c), the Na, K, Cl and SO4 2concentrations are compatible with a natural source. The d34S values of the outcropping evaporites in the area are between +12% and +14%. Therefore, this low value (+8%) could be caused by the addition of isotopically light sulphate from another source, such as fertilisers which may also contribute to the high contents of K. This d34S value can be obtained in a mixture of 75% sulphate from ammonium sulphate fertiliser and natural sulphate (FA–N model) or in a mixture of 95% NPK fertilisers and natural sulphate (FK–N model) (Fig. 6b). Although a natural origin could be assumed for sample 10a from the major ion contents, the Na/K ratio also points to a mining origin. According to the nearby evaporite rocks, the expected sulphur isotope value should be 1371%. However, the measured d34S for that sample is +2.4%, close to the ammonium sulphate fertiliser value which has a sulphur isotopic composition of 1.1%. The calculated mixing models (Fig. 6b, model FA–N) indicate that this isotopic composition can be achieved by an input of B95% of sulphate derived from fertilisers to the waters. Sampling site 9 is located in a small stream with mine tailings nearby. The samples collected at this site (9a and 9b) have d34SSO4=+13.7%; this value is in agreement with the isotopic composition of sulphates from the area (Upper Halite Unit), suggesting a natural origin, which is already supported by the high Na/K ratio of these samples (Fig. 7). This value (around +13%) cannot be explained by a mixture of mining effluents and fertiliser because the Na/K ratio in this case would be an entire order of magnitude lower. The sulphur isotopic composition of dissolved sulphate in waters allows the recognition of the origin of water salinisation in the studied area. The results obtained indicate that the mine effluents are the source
25
20.5
4a 1a
7b 7f 7d 7a 7e 7c
20
17.5
15
δ 34 S (0/00)
3998
9a
3f 3b 3d 3e 3a 3c
3b
2a
6a 6b 6c
9b
6d
5a
10 8a
5 10a
0 0
100
200
300
400
500
600
Na /K
Fig. 7. Plot of Na/K versus d34S for (K) samples of natural origin, (m) samples of mining origin and (’) samples of uncertain origin.
of salinisation at site 7. The samples from site 9 have a salinity of natural origin (dissolution of sulphates from the bedrock) whereas fertilisers are suggested to be major contributors to salinisation for sites 8 and 10.
5. Conclusions This study shows the usefulness of combining elemental and isotopic geochemistry to trace the origin of water salinisation in the middle section of the Llobregat River. The chemical composition of waters allows for the distinction between natural and anthropogenic sources of sulphate. Natural waters are characterised by Na/K ratios higher than 100, whereas waters containing salts from human activities (mining effluents and fertilisers) have ratios lower than 20. The mine tailing effluents are characterised by high Cl, Na, K and Mg content, as well as the occurrence of heavy metals. Groundwaters contaminated by mine effluents have significant metal contents also, therefore, the environmental impact of potash mine tailings is not only water salinisation, but in some cases, the presence of heavy metals as well. According to isotope data three main sources of dissolved sulphate in saline springs can be identified: bedrock evaporites (d34S between +10 and +14%), mining effluents (d34S between +18 and +20%) and fertilisers (d34S=–1.1% for ammonium sulphate and d34S=+7% for NPK). The difference in d34S between the outcropping sulphates and sulphates from the ore unit (potash unit) allows us to know the origin of the
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
sulphate contamination caused by the mining effluents, except in the Cardona Diapir where the potash unit outcrops. Accordingly, this study has shown that the origin of salinisation in the springs located at site 7 is contamination from mine tailing effluents, also a natural origin has been determined for the springs from site 9. Fertilisers used in this area are characterised by sulphur isotopic values of dissolved sulphate lower than the bedrock allowing us to identify their contribution to groundwater salinisation. Elemental and isotope geochemistry point to fertilisers as the major source of sulphate in sites 8 and 10. Coupling isotopic composition analysis together with geochemistry allowed us also to quantify the relative contribution of these sources to groundwater salinisation. Since the data reveal that fertilisers play an important role as salinity source, a more detailed sampling and isotopic characterisation of fertilisers is proposed in a further project. Also, the analysis of the oxygen isotopic composition of sulphate could be used in future studies to assess the extent of the redox processes, which have been considered negligible in the present study.
Acknowledgements This study has been financed by the CICYT project HID99-0498 of Spanish Government, and partially by the SGR99-00062 from the Catalonian Government. We would like to thank J. Capdevila for the pluviometer sampling, Ag"encia Catalana de l’Aigua (Environmental Department of Catalonian Government, especially Llu!ıs Gode! and Josep Bertran) for the field assistance, and the Serveis Cient!ıfics T"ecnics (University of Barcelona). We also wish to thank S.H. Bottrell and an anonymous reviewer for their constructive comments on the paper.
References [1] Cameron EM, Hall GEM, Veizer J, Krouse HR. Isotopic and elemental hydrogeochemistry of a major river system; Fraser River, British Columbia, Canada. Chem Geol 1995;122(1–4):149–69. [2] Hitchon B, Krouse HR. Hydrogeochemistry of the surface waters of the Mackenzie River drainage basin, Canada: stable isotopes of oxygen, carbon and sulfur. Geochim Cosmochim Acta 1972;36:1337–57. . [3] Ingri J, Torssander P, Andersson PS, Morth CM, Kusakabe M. Hydrogeochemistry of sulfur isotopes in the Kalix River catchment, northern Sweden. Appl Geochem 1997;12(4):483–96. [4] Longinelli A, Edmond JM. Isotope geochemistry of the Amazon basin: a reconnaissance. J Geophys Res 1983; 88:3807–17.
3999
[5] Williams MV, Yang D, Liu F, Turk JT, Melack JM. Controls on the major ion chemistry of the Urumqi River, Tian Shan, People’s Republic of China. J Hydrol 1995; 172(1–4):209–29. [6] Yang W, Spencer RJ, Krouse HR. Stable isotope composition of waters and sulfate species therein, Death Valley, California, USA: implications for inflow and sulfate sources, and arid basin climate. Earth Planet Sci Lett 1997;147:69–82. [7] Dowuona GN, Mermut AR, Krouse HR. Stable isotope geochemistry of sulfate in relation to hydrogeology in southern Saskatchewan, Canada. Appl Geochem 1993;8: 255–63. [8] Feast NA, Hiscock KM, Dennis PF, Bottrell SH. Controls on stable isotope profiles in the Chalk aquifer of north-east Norfolk, UK, with special reference to dissolved sulphate. Appl Geochem 1997;12:803–12. [9] van-Donkelaar C, Hutcheon IE, Krouse HR. d34S, d18O, dD in shallow groundwater: tracing anthropogenic sulfate and accompanying groundwater/rock interactions. Water Air Soil Pollut 1995;79:279–98. [10] Bottrell SH, Weber N, Gunn J, Worthington SRH. The geochemistry of sulphur in a mixed allogenic-autogenic karst catchment, Castelton, Derbyshire, UK. Earth Surf Process Landforms 2000;25:155–65. [11] de Caritat P, Krouse HR, Hutcheon I. Sulphur isotope composition of stream water, moss and humus from eight arctic catchments in the Kola Peninsula Region (NW Russia, N Finland, NE Norway). Water Air Soil Pollut 1997;94:191–208. [12] Grinenko VA, Krouse HR, Fedorov YuA. Sulfur-isotope composition as the key to determining whether the salt mass in lake Baykal was formed by natural or man-made processes. Geochem Int 1994;31(1):116–21. [13] Grasby SE, Hutcheon I, Krouse HR. Application of the stable isotope composition of SO4 to tracing anomalous TDS in Nose Creek, southern Alberta, Canada. Appl Geochem 1997;12:567–75. [14] Moncaster SJ, Bottrell SH, Tellam JH, Lloyd JW, Konhauser KO. Migration and attenuation of agrochemical pollutants: insights from isotopic analysis of groundwater sulphate. J Contam Hydrol 2000;43: 147–63. [15] Robinson BW, Bottrell SH. Discrimination of sulfur sources in pristine and polluted New Zealand River catchments using stable isotopes. Appl Geochem 1997;12: 305–19. [16] Yang C, Telmer K, Veizer J. Chemical dynamics of the ‘‘St. Lawrence’’ riverine system; delta D (sub H2O), delta (super 18) O (sub H2O), delta (super 13) C (sub DIC), delta (super 34) S (sub sulfate), and dissolved (super 87) Sr/ (super 86) Sr. Geochim Cosmochim Acta 1996;60(5): 851–66. [17] Mayer B, Fritz P, Prietzel J, Krouse HR. The use of stable sulfur and oxygen isotope ratios for interpreting the mobility of sulfate in aerobic forest soils. Appl Geochem 1995;10(2):161–73. . [18] Morth CM, Torssander P, Kusakabe M, Hultberg H. Sulfur isotope values in a forested catchment over four years: evidence for oxidation and reduction processes. Biogeochemistry 1999;44:51–71.
4000
N. Otero, A. Soler / Water Research 36 (2002) 3989–4000
[19] Soler A, Canals A, Goldstein SL, Otero N, Antich N, Spangerberg J. Sulfur and strontium isotope composition of Llobregat River (NE Spain): tracers of natural and anthropogenic chemicals in stream waters. Water Air Soil Pollut 2002;136:207–24. [20] S!aez A. Estratigraf!ıa y sedimentolog!ıa de las formaciones lacustres del tr!ansito Eoceno—Oligoceno de la Cuenca del Ebro. Translated title: Stratigraphy and sedimentology of the Upper Eocene and Oligocene lacustrine systems of the Ebro Basin. Ph.D. Thesis, University of Barcelona, 1987. 353pp. " [21] Rossell L. Estudi petrologic, sedimentol"ogic i geoqu!ımic de la formacio! de sals pot"assiques de Navarra (Eoc"e sup.). Translated title: Petrology, sedimentology and geochemistry of the Navarra potash formation (Upper Eocene). Ph.D. Thesis, University of Barcelona, 1983. 321pp. [22] God!e Ll. Ag"encia Catalana de l’Aigua, Dept. de Medi Ambient de la Generalitat de Catalunya. C/Provenc-a 204208, 08036, Barcelona, 2001.
[23] Vall"es F. La Salinitzacio! del Riu Llobregat a Sallent. Translated title: Llobregat River salinization at Sallent village. Proceedings of the El ruman sal!ı: impacte sobre el medi natural, urb"a i hum"a de la comarca del Bages, 1998. p. 97–108. [24] Mortvedt JJ. Heavy metal contaminants in inorganic and organic fertilizers. Fert Res 1996;43:55–61. ! D. Evolucion ! geoqu!ımica de cuencas evapor!ıticas [25] Cendon ! isotopica ! terciarias: implicaciones en la composicion del sulfato disuelto en el oc!eano durante el terciario. Translated title: Geochemical evolution of tertiary evaporitic basins: implications for the isotopic composition of dissolved oceanic sulphates during Tertiary Age. Ph.D. Thesis, Universitat de Barcelona, 1999. 270pp. [26] Utrilla R, Pierre C, Ort!ı F, Pueyo JJ. Oxygen and sulphur isotope compositions as indicators of the origin of Mesozoic and Cenozoic evaporites from Spain. Chem Geol (Isot Geosci Sect) 1992;102:229–44.