Journal of Geochemical Exploration 93 (2007) 1 – 12 www.elsevier.com/locate/jgeoexp

Geochemical processes controlling the elevated fluoride concentrations in groundwaters of the Taiyuan Basin, Northern China Qinghai Guo, Yanxin Wang ⁎, Teng Ma, Rui Ma School of Environmental Studies and MOE Biology and Environmental Geology Lab, China University of Geosciences, 430074 Wuhan, China Received 14 February 2006; accepted 6 July 2006 Available online 1 September 2006

Abstract High fluoride groundwater with F− concentration up to 6.20 mg/L occurs in Taiyuan basin, northern China. The high fluoride groundwater zones are mainly located in the discharge areas, especially in places where shallow groundwater occurs (the groundwater depth is less than 4 m). Regional hydrogeochemical investigation indicates that processes including hydrolysis of silicate minerals, cation exchange, and evaporation should be responsible for the increase in average contents of major ions in groundwater from the recharge areas to the discharge areas. The concentration of F− in groundwater is positively correlated with that of HCO−3 and Na+, indicating that groundwater with high HCO−3 and Na+ contents help dissolve some fluoride-rich minerals. The water samples with high F− concentration generally have relatively higher pH value, implying that alkaline environment favors the replacement of exchangeable F− in fluoride-rich minerals by OH− in groundwater. In addition, the mixing of karst water along the western mountain front and the evaporation may also be important factors for the occurrence of high fluoride groundwater. The inverse geochemical modeling using PHREEQC supports the results of hydrogeochemical analyses. The modeling results show that in the recharge and flow-through area of the northern Taiyuan basin, interactions between groundwater and fluoride-rich minerals are the major factor for the increase of F− concentration, whereas in the discharge area of the northern basin, the evaporation as well as the mixing of karst water has greater contribution to the fluoride enrichment in groundwater. © 2006 Elsevier B.V. All rights reserved. Keywords: Fluoride; Groundwater; Geochemistry; Water–rock interaction; Mixing; Evaporation; Taiyuan basin; China

1. Introduction Fluorine has the highest chemical reactivity among all known elements and occurs mainly as free fluoride ions in natural waters, although some fluoride complexes also exist under specific conditions. It was reported that the assimilation of fluoride by the human body from drinking water with concentrations above 1.5 mg/L might result in fluorosis (Rukah and Alsokhny, 2004; Ghorai and Pant, 2005). However, other studies indicate that the occurrence ⁎ Corresponding author. Tel.: +86 27 62879198; fax: +86 27 87481030. E-mail address: [email protected] (Y. Wang). 0375-6742/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gexplo.2006.07.001

of fluorosis is dependent on a number of factors. For example, the molarity ratio of F to Ca in groundwater has evident effect on fluorosis. It had been also found out that Mg and P in groundwater might restrain the assimilation of fluoride by the human body (Shen et al., 1999). But generally speaking, in areas using high fluoride groundwater for water supply, endemic fluorosis may occur. Endemic fluorosis develops widely in many areas of the world, such as Mexico (Grimaldo et al., 1995), East Africa (Gaciri and Davies, 1993), India (Jacks et al., 2005), and China (Wang and Huang, 1995; Cao et al., 2001; Dai et al., 2004). The causes not only include longterm intake of high fluoride groundwater, but exposure to

2

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

high fluoride gas from coal burning as well (Wang and Huang, 1995). In addition, it was reported that drinking brick tea in excess is also an important cause for endemic fluorosis in western China where brick tea is produced and consumed (Cao et al., 1997, 2000, 2003). For the fluoride in groundwater, common natural sources are the dissolution of following minerals, such as fluorspar, fluorapatite, amphiboles (e.g., hornblende, tremolite) and some micas weathered from silicates, igneous and sedimentary rocks, especially shale (Datta et al., 1996). Unstable minerals such as sepiolite and palygorskite may have a dominant control on F− distribution in groundwater as well (Wang et al., 1993; Jacks et al., 2005). Since drinking high fluoride groundwater is the major reason for endemic fluorosis and has considerable impact on human health, many efforts have been made in recent years to study the hydrochemistry and genesis of high fluoride groundwater as well as alternative technologies of defluoridation (Handa, 1975; Corbett and Manner, 1984; Sarma and Rao, 1997; Gosselin et al., 1999; Wang and Reardon, 2001; Rukah and Alsokhny, 2004).

Taiyuan basin is a representative large-scale Cenozoic rift basin in Shanxi Province, northern China. Groundwater has been the most important source of water supply. However, groundwaters with high fluoride concentration above 1.5 mg/L were found in some parts of the basin. To delineate high fluoride groundwater zones and to understand the basic hydrogeological and geochemical processes controlling fluoride enrichment are extremely important for water resource management and environmental protection in this semi-arid region. 2. Regional hydrogeology Taiyuan basin is located in the central part of Shanxi Province, with an area of 6159 km2. The annual mean air temperature of the basin has varied from 8.8 °C to 10.6 °C in the past 50 years. The average annual rainfall from 1951 to 2000 was 446.6 mm, the largest was 655.0 mm (1964) and the smallest was 259.8 mm (1972). The annual mean potential water surface

Fig. 1. (A) Simplified geological map of Taiyuan Basin; (B) Hydrogeological cross section along the I–II line.

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

evaporation is 1774.9 mm. Surface waters in the study area are mainly Fen River and its tributaries. Bedrock around the Taiyuan basin includes Cambrian–Ordovician carbonate rocks, Carboniferous–Permian coal-bearing strata and Triassic clastic rocks (Fig. 1). The thickness of Cenozoic sediments ranges between 50 m

3

and 3800 m. The grain size of the sediments generally decreases from the margin to the center of the basin. The Quaternary groundwater system can be vertically divided into two groups that are separated by a set of clay layer: the Holocene and the upper–middle Pleistocene aquifer with the buried depth of 0–200 m,

Fig. 2. Sampling locations and F− content contours in Taiyuan basin. The Basin can be divided into northern part and southern part by Tianzhuang fault. Zone A and B are respectively the recharge and flow-through area and the discharge area in the northern part, and Zone C and D respectively are those in the southern basin.

4

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

Table 1 Average hydrochemistry of groundwater samples in Zones A, B, C and D (in mg/L except pH. Figure in parentheses is the total number of groundwater samples) Average hydrochemistry

pH

HCO−3

SO2− 4

Cl−

NO−3

Si

Ca2+

Mg2+

Na+

K+

F−

TDS

Zone A (22) Zone B (4) Zone C (20) Zone D (12)

7.58 8.08 7.07 7.85

316 501 351 579

166 1008 297 407

84.5 168 104 435

8.6 29.0 17.8 52.3

5.8 7.2 6.4 7.3

102 227 104 121

44.4 106 48.6 89.4

57.4 258 147 381

0.89 1.47 6.18 3.18

0.59 1.97 0.91 1.78

622 2049 900 1780

and the lower Pleistocene aquifer with the buried depth of 200–400 m. For the upper Quaternary aquifers, the sediments from the basin margin are mostly alluvial– pluvial gravel and sand, whereas those at the central part of the basin are silty sand and silty clay; for the lower Quaternary aquifers, the sediments are mainly composed of lacustrine and alluvial–lacustrine sandy loam, silty clay and clay. In this study, groundwater was sampled from the Holocene and the upper–middle Pleistocene aquifer, the most important water supply source in the basin. With Tianzhuang fault being the regional groundwater divide, Taiyuan Basin can be divided into northern part and southern part (Fig. 2). In the northern part, groundwater flows towards the discharge area (Zone B, Fig. 2) from the north, west, and east of the recharge and flow-through area (Zone A, Fig. 2), whereas in the southern part, the discharge area (Zone D, Fig. 2) is surrounded by the recharge and flow-through area (Zone C, Fig. 2). Generally, groundwater recharge mainly occurs in two ways: via vertical seepage of meteoric water in the basin and via lateral penetration of karst water and fissure water along the mountain front. In addition, leakage from local rivers and irrigation return flow should also be taken into account as groundwater recharge sources. The major ways of discharge include evaporation and artificial abstraction.

aquifer were collected between May and September 2003, and the locations of samples are shown in Fig. 2. The average hydrochemistry of groundwater samples in Zones A, B, C, and D are listed in Table 1. When sampling, all water samples were filtered through 0.45 μm membranes on site. Samples were collected in three new 350 ml polyethylene bottles. For metallic element analysis, reagent-quality HNO3 with molar concentration of up to 14 M was added to one of these polyethylene bottles until pH of samples reached 1. These bottles had been rinsed with deionized water twice before sampling. Unstable hydrochemical parameters including water temperature and pH were measured in-situ using portable Hanna EC and pH meter that had been calibrated before use. Alkalinity was measured on the sampling day using the Gran titration method (Appelo and Postma, 1996). F− and major anions and cations (HCO3− , SO42− , Cl− , NO3− , Ca2+ , Mg2+ , Na+ , K + ) were determined using ultraviolet spectrophotometer and ion chromatography respectively, and SiO2 by colorimetry within 2 weeks after sampling. Five sediment samples were collected from the basin, and their mineralogy determined using X-ray powder diffractometry (XRD) are given in Table 2.

3. Methods

Inverse geochemical modeling can determine the nature and extent of the geochemical reactions by identifying the reacting minerals and the dissolution or precipitation amount of these minerals (Plummer and Back, 1980; Hidalgo and Cruz-Sanjulián, 2001). The geochemical computer code PHREEQC capable of

3.1. Sampling and analysis Fifty-nine water samples including 58 from sedimentary aquifer plus 1 karst water sample from carbonate

3.2. Inverse geochemical modeling

Table 2 Mineral compositions of sediment samples estimated from XRD analysis results No.

Lithology

Quartz

Feldspar

Micas

Calcite

Kaolinite

Chlorite

Amphibole

13 15 27 32 28

19 17 15 8 10

9 12 10 10 12

30 15 – – –

8 18 15 10 9

– – 7 15 12

(%) TB01 TB02 TB03 TB04 TB05

Clay Silty clay Silty sand Fine sand Coarse sand

21 23 26 25 29

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

calculating mass transfer along groundwater flow paths (Parkhurst, 1997) was used in this study. PHREEQC 2.6 is an interactive C computer program that can be used to compute the mass transfer between groundwater and minerals based on the observed compositions of two consecutive water samples along certain

5

groundwater flow path. In the present study, this software was run to calculate the dissolution/precipitation amount of the major aquifer minerals along a groundwater flow path in the northern Taiyuan basin and evaluate the effect of major hydrogeochemical processes on the F− concentration in groundwater. It is important to stress that the

Fig. 3. Box and whisker plots of the concentrations of major ions in groundwater samples in Zones A, B, C and D. The boxes show the mean value minus standard error, the mean value, and the mean value plus standard error. The smallest and largest values are indicated by the small horizontal bars at the end of the whiskers.

6

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

geochemical reaction model built in this study may be just one of the descriptions for the actual states. The solutions produced by PHREEQC are critically dependent on the constraints, phases and assumptions included in the model. So the results given in the paper are not the only explanation of the groundwater compositions. 4. Results and discussion 4.1. Groundwater chemistry At Taiyuan, the pH values of groundwater samples are between 6.00 and 8.80, and the average value is 7.49. The TDS values of most groundwater samples are less than 1.5 g/L, the maximum being 8.01 g/L. The main cations of groundwaters are Na+, Ca2+, and Mg2+, and the main anions are HCO3−, SO42−, and Cl−. Regional groundwater chemistry changes in different parts of the basin, and in both northern and southern part, the average contents of HCO3−, SO42−, Cl−, Ca2+, Mg2+, Na+ and F− in groundwater all increase from the recharge and flowthrough area (Zones A and C) to the discharge area (Zones B and D), as shown in Fig. 3. The spatial distribution of F− content is shown in Fig. 2, where it can be seen that the F− content of groundwater generally increases along the groundwater flow paths. In Taiyuan basin, high fluoride groundwater is mostly found in the discharge areas (Zone B and Zone D), although a few samples with high F− content occur in the mountain front, such as well 30 in Zone A and wells 3, 4 in Zone C.

Fig. 5. [Na]/[Ca] ratio versus TDS value of groundwater samples in Zones A, B, C, D.

According to the mineral analysis results in Table 2, quartz, feldspar and micas are the major minerals of Quaternary sediments, and the hydrolysis of feldspar and micas should be an important geochemical process controlling the groundwater chemistry at Taiyuan. In the mineral stability diagram of the Na+–H+–SiO2 system (Fig. 4), all the groundwater samples fall within the kaolinite stability field, indicating that some silicate minerals, such as plagioclase and biotite, may have been hydrolyzed to kaolinite. The dissolution of these minerals can produce kaolinite plus cations such as Ca2+, Mg2+, Na+ and K+, resulting in their concentration increase along the groundwater flow paths. The hydrolysis reaction equations are as follows: Plagioclase: Na0:62 Ca0:38 Al1:38 Si2:62 O8 þ 1:38CO2 þ 4:55H2 O ¼ 0:69Al2 Si2 O5 ðOHÞ4 þ 0:62Naþ þ 0:38Ca2þ þ 1:24H4 SiO4 þ 1:38HCO−3

Biotite: KMg3 AlSi3 O10 ðOHÞ2 þ 7CO2 þ 7:5H2 O ¼ 0:5Al2 Si2 O5 ðOHÞ4 þ Kþ þ 3Mg2þ þ 2H4 SiO4 þ 7HCO−3

Fig. 4. All of the groundwater samples plotted onto the Na+–H+–SiO2 system equilibrium phase diagram. Qtz. Sat. refers to quartz saturation line, and Am. SiO2 saturation amorphous SiO2 saturation line.

ð1Þ

ð2Þ

Although the contents of Ca2+ and Na+ in groundwater all increase from the recharge and flow-through area to the discharge area, the increase of the Na+ is more evident compared with that of Ca2+. The [Na]/[Ca] ratio of groundwater in the discharge areas, especially that in

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

7

Fig. 6. Piper diagram of groundwater samples collected from the discharge areas as delineated on Fig. 2.

Zone D, is much higher than that in the recharge and flowthrough area (Fig. 5). The Piper diagram (Fig. 6) also shows that most samples in the discharge areas contain Na+ as the predominant cation. The high Na+ concentration in groundwater of the discharge areas may be related to the cation exchange occurring in the aquifers. In the margin of the basin, the aquifer media are mainly composed of coarse-grained sediment (its major minerals are quartz, feldspar, etc.) with limited ion exchange capacity, and as a result the [Na]/[Ca] ratio is comparatively lower. But in the discharge areas, the aquifers contain more fine sediments, especially clay minerals that can take up Ca2+ from groundwater and release Na+ into groundwater, resulting in an increase in Na+ concentration, a decrease in Ca2+ content in groundwater, and consequently an increase in [Na]/[Ca] ratio. It is quite interesting to note that the case studies of Guo and Wang (2005) in another Cenozoic rift basin of northern China, Datong basin where groundwaters with high F− concentration occur also showed that the high [Na]/[Ca] ratio in the stagnant groundwater environment has close relation with the ion exchange between Na+ absorbed on the surface of clay minerals and Ca2+ in groundwater. In the discharge area, some samples have abnormally high TDS values, such as well 40 (TDS value = 5.02 g/L)

in Zone B and well 1 (TDS value = 8.01 g/L) in Zone D. Since these samples are just located in areas where shallow groundwaters occur (namely the groundwater depth is less than 4 m, Fig. 2), it can be inferred that the evaporation process be responsible for the high TDS value of these samples. However, some samples collected in the mountain front have high TDS value as well, such as well 4 in Zone C (TDS value = 4.22 g/L). A further inspection of the results indicates that wells 30, 37 in Zone A, wells 40, 41 in Zone B and wells 3, 4, and 51 in Zone C from the western mountain front all have high SO42− content, with an average value of 1210 mg/L. From Fig. 2, it can be seen that there is a large-scale coal mining area in the west to the basin. The wastewater from the mine Table 3 Average hydrochemistry of high-fluoride groundwater with F− content more than or equal to 1.5 mg/L and low-fluoride groundwater with F− content less than 1.5 mg/L (in mg/L except pH. Figure in parentheses is the total number of groundwater samples) Average hydrochemistry −

F−

pH

Na+ HCO−3 TDS

Groundwater with F content more 2.45 7.84 364 530 than or equal to 1.5 mg/L (14) Groundwater with F− content less 0.59 7.38 107 352 than 1.5 mg/L (44)

2221 686

8

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

concentration and TDS value of the wastewater are 2446 mg/L and 3850 mg/L respectively, so its discharge would increase the SO42− concentration of karst waters remarkably. The SO42− concentration of karst water sample from well 59 in the coalmine area (Fig. 2) is up to 1126 mg/L, indicating the impact of mining wastewater. As mentioned above, the lateral permeation of karst water is the major way of groundwater recharge at Taiyuan, and it can therefore be postulated that the mixing of karst water added a great amount of SO42− into groundwater and is an important factor controlling groundwater chemistry along the western mountain front. 4.2. Genesis of high fluoride groundwater

Fig. 7. (A) F− versus HCO−3 of all groundwater samples; (B) F− versus Na+ of all groundwater samples. The trend lines in (A) and (B) show the linear correlation relationships with correlation coefficient R of 0.650 and 0.785 respectively. Note that samples from wells 3 and 4 in Zone C downstream of a large-scale coalmine deviate from the trend line remarkably.

directly drains out without treatment and then infiltrates where there are depressions with limestone outcrops. According to the work by Geological Survey of Shanxi Province in 2004 on the effect of coal mining on groundwater quality in Taiyuan basin, the average SO42−

To investigate the genesis of high fluoride groundwater, the hydrochemistry of water samples with F− content higher than or equal to 1.5 mg/L was analyzed, the results indicating that the high fluoride groundwater typically has high pH value (the average and maximum value being 7.84 and 8.60 respectively), high HCO3− content (the average and maximum value being 530 and 934 mg/L respectively), and high Na+ content (the average and maximum value being 364 and 980 mg/L respectively). Moreover, the comparison between the high fluoride water samples with F− content more than or equal to 1.5 mg/L and the low fluoride water samples with F− content less than 1.5 mg/L shows that the average pH value, HCO3−, Na+ and TDS value of the former are all much higher than those of the latter (Table 3). The plots of F− versus HCO3− and F− versus Na+ (Fig. 7) also indicate that fluoride concentration is positively correlated with HCO3− and Na+ contents, with the regression coefficient of 0.650 and 0.785 respectively. For F− in groundwater, fluorite may be a natural mineral source, and F− is also abundant in some hydroxy-minerals such as muscovite, biotite, and apatite (Jacks et al., 2005). The weathering of these minerals is the source of F− for the subsurface environment. However, the F− in groundwater is mainly derived from the leachable or exchangeable F− that only accounts for a small proportion of the total fluorine in aquifer materials

Table 4 Hydrochemical properties and major ions of water samples along the groundwater flow path (as delineated on Fig. 2) and karst water sample (in mg/L except pH; temperature as T in °C) Well

pH

T

Ca2+

Mg2+

Na+

K+

HCO−3

SO2− 4

Cl−

F−

Si

49 28 36 40 59

7.10 7.20 7.70 8.30 6.41

13.3 14.2 18.2 15.6 24.5

56.0 70.1 131 534 322

32.1 35.2 41.3 205 97.1

56.3 76.2 97.0 646 43.8

0.30 0.60 0.98 2.36 0.10

175 218 232 870 317

134 142 339 2654 1126

75.3 96.3 117 434 28.1

0.40 0.52 0.71 3.32 1.89

6.0 6.7 8.6 8.9 5.9

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12 Table 5 Constraints and phases used in the PHREEQC model Elements as constraints Silicon

Phases

Calcium

Plagioclase Kaolinite Ca–Na exchange Carbon Magnesium Biotite Gypsum OH–F (inorganic carbon) exchange Sulfur Sodium Quartz Halite Chlorine Potassium Calcite Fluorite Fluorine CO2 (g)

(Zhang et al., 1998; Zhu and Yin, 2005). For the dissolution of fluorite in groundwater with high HCO3− contents, the reaction is as follows: CaF2 þ 2HCO−3 ¼ CaCO3 þ 2F− þ H2 O þ CO2

ð3Þ

Moreover, groundwaters with high HCO3− and Na+ content are usually alkaline and have relative high OH− content, so the OH− can replace the exchangeable F− of fluoride-bearing minerals, increasing the F− content in groundwater. The reactions are basically as follows: Muscovite: KAl2 ½AlSi3 O10 F2 þ 2OH− ¼ KAl2 ½AlSi3 O10 ½OH2 þ 2F−

ð4Þ

Biotite: KMg3 ½AlSi3 O10 F2 þ 2OH− ¼ KMg½AlSi3 O10 ½OH2 þ 2F−

ð5Þ

Thus, as indicated by regional hydrochemical data, groundwater with high HCO3− and Na+ contents occurs in the discharge areas (Zone B and Zone D) at Taiyuan as a result of silicate mineral hydrolysis and cation exchange. So the occurrence of groundwater with high HCO3− and Na+ contents and high pH value under the control of above water–rock interactions is the important reason for fluoride release from the aquifer matrix into groundwater. In addition to interaction between groundwater and fluoride-rich minerals, evaporation is another important factor resulting in the occurrence of high fluoride groundwater. As a result of evaporation, Ca2+ would

9

precipitate out as CaCO3, reducing Ca2+ concentration of the groundwater, and consequently the solubility control of CaF2 on fluoride enrichment in the aqueous phase becomes weaker. In Taiyuan basin, most water samples in the evaporation zone are over saturated with calcite due to evaporation and the calcite precipitation can reduce the Ca2+ content and promote the dissolution of CaF2. The same relationship between evaporation and fluoride enrichment was noticed by Datta et al. (1996) at Rajasthan, India, where strong evapotranspiration is also associated with the high F− content in the groundwater. As mentioned above, well 1 in Zone D and well 40 in Zone B have very high TDS values (8.01 and 5.02 g/L respectively) and were collected in areas where the groundwater depth is less than 4 m, implying that the hydrochemistry of these two samples is closely related to evaporation. Interestingly, their fluoride concentrations are quite high, up to 3.65 and 3.32 mg/L respectively. Moreover, it is noticeable that on the F− versus HCO3− and F− versus Na+ plot (Fig. 7), although most groundwater samples cluster around the trend line, some samples with high F− content deviate from it remarkably, such as wells 3 and 4 in Zone C. These two samples collected from the western mountain front have high SO42− contents (696 and 2486 mg/L respectively). Therefore, mixing of karst water should be one of the factors responsible for the deviation. However, the F− concentration of well 3 in Zone C (6.20 mg/L) is much higher than that of karst water sample from well 59 (1.89 mg/L), implying that there are additional sources of F− for well 3 in Zone C, such as the wastewater drainage from local glass or brick making industries. 4.3. Geochemical modeling To evaluate the effect of above-mentioned hydrogeochemical processes on the F− concentration in groundwater, an inverse geochemical model was built using available hydrochemical data. According to the groundwater flow directions as designated in Fig. 2, a groundwater flow path in the northern basin was discerned (well 49 in Zone A – well 28 in Zone A – well 36 in Zone A – well 40 in Zone B), and the hydrochemistry

Table 6 Results of mass transfer along the groundwater flow path in the northern part of the basin Phase

Plagioclase Biotite Quartz

Well 49–28 0.2253 Well 28–36 0.1228 Well 36–40 0.0522

Calcite CO2 (g)

Kaolinite Gypsum Halite

Fluorite Ca– Na

OH–F Mixing ratio Evaporation of karst water coefficient

0.0077 − 0.2830 0.1680 0.5071 −0.1593 0.0778 0.5929 0.0032 – – – 0.0097 − 0.1400 – 0.1545 −0.0896 1.6301 0.5743 0.0050 – – – 0.0002 − 0.0674 – − 0.1011 − 0.0379 1.4420 1.5180 – 2.1010 0.0037 45%

All mineral and gas mass transfers are in mmol/kg H2O.

– – 4

10

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

of the above samples can be seen in Table 4. The constraints and phases used in the PHREEQC modeling were selected (Table 5) based on potential hydrogeochemical reactions taking place in Quaternary aquifers. As the hydrolysis of silicate minerals, such as plagioclase and biotite, is the major water–rock interaction, these minerals can act as phases in the model, and the corresponding elements Si, Ca, Mg, Na, K as constraints. Because all the samples along groundwater flow path are over-saturated with respect to quartz according to our saturation index calculation, it was set to the status of “only precipitate” along the groundwater flow path. Moreover, calcite as one of the main minerals of sediments (Table 2) was considered as phase and the element C as constraint. The dissolution of gypsum, halite and fluorite most likely provides inputs of SO42−, Cl− and F− to the groundwater, and therefore they were added in the model as phases and the elements S, Cl, F as the corresponding constraints. In addition, Ca–Na exchange and OH–F exchange were included in the model to simulate the high Na+ and F− concentration of groundwater in the discharge area. When the hydrogeochemical reactions in the flow path from well 36 in Zone A to well 40 in Zone B were modeled, the mixing of karst water and the evaporation in the discharge area were taken into account as well, and the sample from well 59 was selected as the representative of karst water. The sample from well 49 in Zone A was used as initial water, from which the mass transfer between groundwater and minerals along the flow path was calculated. The results in every two consecutive water samples were shown in Table 6 where negative values indicate precipitation between initial and final water, and positive indicate dissolution. Moreover, for Ca–Na exchange, negative values indicate that Ca2+ enters groundwater and Na+ is absorbed by aquifer media, and positive indicate that Na+ enters groundwater and Ca2+ is absorbed by aquifer media. For OH–F exchange, negative values indicate that OH− enters groundwater and F− is absorbed by aquifer media, and positive indicate that F− enters groundwater and OH− is absorbed by aquifer media. The mixing ratio of 45% means that the mixing percentage of karst water and groundwater in the discharge area are respectively 45% and 55%. The evaporation coefficient of 4 means that groundwater in the discharge area was concentrated by 4 fold to generate the final water solution. In the flow path from well 49 to well 28, the hydrolysis of plagioclase and biotite is the major geochemical process. Moreover, as can be seen in Table 6, the hydrochemical composition of well 28 in Zone A was also affected by the input of calcite (0.1680 mmol/kg H2O, namely the dis-

solution amount of calcite is 0.1680 mmol per 1 kg H2O, the similar below) and halite (0.5929 mmol/kg H2O), along with a small quantity of gypsum (0.0778 mmol/kg H2O) and fluorite (0.0032 mmol/kg H2O) dissolution. From well 28 to well 36, the reactive state (dissolution or precipitation) of all minerals except for calcite is the same to those from well 49 to well 28, but the minerals' dissolution/ precipitation amount changes evidently. It is worth noticing that the dissolution of fluorite (0.005 mmol/kg H2O) is yet the controlling factor increasing the F− concentration in groundwater, which is similar to the state along the path from well 49 to well 28. The next well 40 lies in the discharge area of the northern basin (Zone B) and represents the groundwater chemistry that was affected by the mixing of karst water and the evaporation. In the flow path from well 36 to well 40, the dissolution amount of plagioclase and biotite (0.0522 mmol/kg H2O and 0.0002 mmol/kg H2O respectively) are evidently smaller than those from well 49 to well 28 and from well 28 to well 36, for groundwaters in the discharge area have been close to saturation for these minerals. The Ca–Na exchange (2.1010 mmol/kg H2O, namely 2.1010 mmol Ca2+ is absorbed by aquifer media from 1 kg H2O, and 4.2020 mmol Na+ enters 1 kg H2O from aquifer media) instead of the hydrolysis of silicate minerals was responsible for the sharp increase of Na+ concentration in groundwater. With the increase of the pH value of groundwater in the discharge area (average pH value 8.08), OH–F exchange (0.0037 mmol/kg H2O) became one of the factors controlling the F− concentration in groundwater. Furthermore, the mixing ratio of karst water and the evaporation coefficient of groundwater were calculated to be 45% and 4 respectively. These two hydrochemical processes change the groundwater chemistry evidently, especially the F− concentration. From well 36 to well 40, the increase of F− concentration in groundwater as a result of water–minerals interactions, that of mixing of karst water, and that of evaporation are respectively 0.14 mg/L, 0.37 mg/L, and 2.10 mg/L. In other word, the contributions of the above three processes to fluoride enrichment in groundwater are respectively 5%, 15% and 80%. Thus it can be seen that the evaporation and the mixing of karst water are more important than water–minerals interactions in controlling fluoride chemistry of groundwater in the discharge area of the northern part at Taiyuan. It is worth noting that the above solutions obtained using PHREEQC is critically dependent on the minerals, constraints and assumptions included in the inverse geochemical model, namely the results in this paper are not unique solutions. Other explanations of the groundwater

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

11

compositions may be acquired in case that the different constraints and phases are used in inverse modeling.

fited from the helpful comments from two anonymous reviewers.

5. Conclusions

References

From the recharge and flow-through areas to the discharge areas of Taiyuan basin, the average contents of HCO3−, SO42−, Cl−, Ca2+, Mg2+, Na+ and F− in groundwater all increase, as a result of hydrolysis of silicate minerals along groundwater flow paths, together with cation exchange and evapotranspiration in the discharge areas. The high fluoride groundwater zones are located in the discharge areas, especially the places where shallow groundwaters occur (the groundwater depth is less than 4 m). The high fluoride groundwaters with F− content more than or equal to 1.5 mg/L are characterized by relative higher pH value, HCO3− and Na+ contents, as compared with the low-fluoride groundwaters with F− content less than 1.5 mg/L. Correlation analysis also indicates that the fluoride content in groundwater is positively correlated with the HCO3− and Na+ contents. This implies that alkaline condition is favorable for the substitution of exchangeable F− in fluoride-rich minerals by OH− in groundwater. Besides, groundwaters with high HCO3− and Na+ contents have greater potential to dissolve fluoride-rich minerals, such as fluorite, under the control of CaF2 solubility. The mixing of karst water along the western mountain front and the evaporation under the semi-arid conditions at Taiyuan are the important factors controlling the occurrence of high fluoride groundwater as well. The results of inverse geochemical modeling using PHREEQC also indicate that in the recharge and flow-through area in the northern part of the basin, the interactions between groundwater and fluoride-rich minerals in the aquifers are responsible for the increase of F− concentration in groundwater along the flow path, whereas in the discharge area, the evaporation and the mixing of karst water with high F− concentration have more significant effect on the fluoride enrichment in groundwater than water–rock interactions.

Appelo, C.A.J., Postma, D., 1996. Geochemistry, groundwater, and pollution. Balkema, Rotterdam. Cao, J., Zhao, Y., Liu, J., 1997. Brick tea consumption as the cause of dental fluorosis among children from Mongol, Kazak and Yugu populations in China. Food and Chemical Toxicology 35, 827–833. Cao, J., Zhao, Y., Liu, J., 2000. Fluoride in the environment and bricktea-type fluorosis in Tibet. Journal of Fluorine Chemistry 106, 93–97. Cao, J., Zhao, Y., Liu, J., Xirao, R., Danzeng, S., 2001. Varied ecological environment and fluorosis in Tibetan children in the nature reserve of Mount Qomolangma. Ecotoxicology and Environmental Safety 48, 62–65. Cao, J., Zhao, Y., Liu, J., Xirao, R., Danzeng, S., Daji, D., Yan, Y., 2003. Brick tea fluoride as a main source of adult fluorosis. Food and Chemical Toxicology 41, 535–542. Corbett, R.G., Manner, B.M., 1984. Fluoride in the ground water of northeastern Ohio. Ground Water 22 (1), 13–17. Dai, S., Ren, D., Ma, S., 2004. The cause of endemic fluorosis in western Guizhou Province, Southwest China. Fuel 83, 2095–2098. Datta, P.S., Deb, D.L., Tyagi, S.K., 1996. Stable isotope (18O) investigations on the processes controlling fluoride contamination of groundwater. Journal of Contaminant Hydrology 24, 85–96. Gaciri, S.J., Davies, T.C., 1993. The occurrence and geochemistry of fluoride in some natural waters of Kenya. Journal of Hydrology 143, 395–412. Ghorai, S., Pant, K.K., 2005. Equilibrium, kinetics and breakthrough studies for adsorption of fluoride on activated alumina. Separation and Purification Technology 42, 265–271. Gosselin, D.C., Headrick, J., Harvey, F.E., Tremblay, R., McFarland, K., 1999. Fluoride in Nebraska's ground water, vol. 19 (2), pp. 87–95. Grimaldo, M., Borja-Aburto, V.H., Ramirez, A.L., Ponce, M., Rosas, M., Diaz-Barrjga, F., 1995. Endemic fluorosis in San Luis Potosi, Mexico. 1. Identification of risk factors associated with human exposure to fluoride. Environmental Research 68, 25–30. Guo, H., Wang, Y., 2005. Geochemical characteristics of shallow groundwater in Datong basin, northwestern China. Journal of Geochemical Exploration 87, 109–120. Handa, B.K., 1975. Geochemistry and genesis of fluoride-contamination ground waters in India. Ground Water 13 (3), 275–281. Hidalgo, M.C., Cruz-Sanjulián, J., 2001. Groundwater composition, hydrochemical evolution and mass transfer in a regional detrital aquifer. Applied Geochemistry 16, 745–758. Jacks, G., Bhattacharya, P., Chaudhary, V., Singh, K.P., 2005. Controls on the genesis of some high-fluoride groundwater in India. Applied Geochemistry 20, 221–228. Parkhurst, D.L., 1997. Geochemical mole-balance modeling with uncertain data. Water Resources Research 33, 1957–1970. Plummer, L.N., Back, W.W., 1980. The mass balance approach— application to interpreting the chemical evolution of hydrologic systems. American Journal of Science 280, 130–142. Rukah, Y.A., Alsokhny, K., 2004. Geochemical assessment of groundwater contamination with special emphasis on fluoride concentration, North Jordan. Chemie der Erde Geochemistry 64, 171–181. Sarma, D.R.R., Rao, S.L.N., 1997. Fluoride concentrations in ground waters of Visakhapatnam, India. Bulletin of Environmental Contamination and Toxicology 58 (2), 241.

Acknowledgements The research work was financially supported by the National Natural Science Foundation of China (40425001), the Ministry of Science and Technology of China (2004DFB01600), the Ministry of Education of China (105033) and the Research Foundation for Outstanding Young Teachers, China University of Geosciences (Wuhan) (CUGQNL0619). The manuscript greatly bene-

12

Q. Guo et al. / Journal of Geochemical Exploration 93 (2007) 1–12

Shen, Z., Zhu, Y., Zhong, Z., 1999. Introduction to Hydrogeochemistry. Geological Press, Beijing. In Chinese. Wang, L., Huang, J., 1995. Outline of control practice of endemic fluorosis in China. Social Science and Medicine 41 (8), 1191–1195. Wang, Y., Reardon, E.J., 2001. Activation and regeneration of a soil sorbent for defluoridation of drinking water. Applied Geochemistry 16, 531–539. Wang, Y., Nahon, D., Merino, E., 1993. Geochemistry and dynamics of calcrete genesis in semi-arid regions. Chemical Geology 107, 349–351.

Zhang, J., Zhu, F., Chen, R., Miao, X., Yao, S., 1998. Inquiry on the genesis of groundwater with high fluoride in Fengxian, Peixian, and Tongshan counties of Jiangsu Province. Geological Journal of China Universities 4 (2), 140–146 in Chinese with English abstract. Zhu, L., Yin, G., 2005. Edaphic fluorine in high section of groundwater of Jiaozuo area. Journal of Henan Polytechnic University 24 (5), 366–368 in Chinese with English abstract.