HYDROCHEMISTRY OF WATERS OF VOLCANIC ROCKS: THE … · 2011-02-24 · HYDROCHEMISTRY OF WATERS OF VOLCANIC ROCKS: THE CASE OF THE VOLCANOSEDIMENTARY ROCKS OF THRACE, GREECE C. PETALAS1,∗,N.LAMBRAKIS2

HYDROCHEMISTRY OF WATERS OF VOLCANIC ROCKS: THE CASEOF THE VOLCANOSEDIMENTARY ROCKS OF THRACE, GREECE

C. PETALAS1,∗, N. LAMBRAKIS2 and E. ZAGGANA3

1Laboratory of Ecological Engineering and Technology, Department of Environmental Engineering,Democritus University of Thrace, Vas. Sofias 12, 67100 Xanthi, Greece; 2University of Patras,

Department of Geology, Laboratory of Hydrogeology, 26500 Patras, e-mail:[email protected]; 3Hellenic Centre for Marine Research (HCMR), Institute of Oceanography,

Athens-Sounio Avenue, 19 013 Anavyssos, Greece, e-mail address: [email protected](∗author for correspondence, e-mail: [email protected], Tel: +30-541-0-79385,

Fax: +30-541-0-79385)

(Received 21 February 2005; accepted 2 September 2005)

Abstract. This work is referred to the characterization of the environmental hydrochemistry in thebroader Sapes area – Thrace region, on the basis of physico-chemical properties of surface andgroundwaters occurring in the volcanosedimentary formations of this area, where gold mining activ-ities are planned to operate. Volcanic rocks are considerably altered where they are in contact withhydrothermal solutions. Aquifers are formed within these formations. Surface and ground waters arestrongly metalliferous and their hydrochemical facies present similar but complex water types. Cer-tain characteristic chemical types are the following: Ca-Mg-HCO3-SO4, Ca-Mg-SO4-HCO3. Ca-SO4,Ca-Mg-SO4. Ca-Na-Cl-HCO3, Na-Cl. A small majority of the water samples present the followingorder of anion dominance HCO3

− > SO42− > Cl−. Calcium is the dominant cation. Bicarbonates

and sulfate ions are the dominant anions. The order of dominance for the heavy metals in surfaceand ground waters is as follows: Fe > Mn > Zn > Ni > Cu. The saturation index of waters regard-ing minerals is low. Computer simulation indicates that calcite and dolomite are common mineralsin all water samples which are saturated in respect to quartz and argillaceous-siliceous minerals.The most pronounced property of waters is their acidic character. The high metal concentrationsare related to water with low pH. Sulfide minerals control the low pH values of waters which isan important control factor for the evolution of the water chemical composition. The abundance ofsulfates is attributed to the dissolution of the minerals pyrite (FeS2) and alunite (KAl3(SO4)2(OH)6.The water–mineral interactions are responsible for the chemical composition of waters. Water qual-ity problems can be successfully handled by the use factor analysis. 17 chemical parameters canbe substituted by five factors which successfully represent the hydrochemical processes as well astheir geographic distribution. Volcanic rocks in the study area have the potential to produce aciddrainage.

Keywords: Acid water, groundwater, hydrochemistry, saline water, volcanic rocks

1. Introduction

Recently, environmental problems caused by acid drainage, are evident in manyparts of the world. In the last few years, these problems have been intensely studied

Water, Air, and Soil Pollution (2006) 169: 375–394 C© Springer 2006

376 C. PETALAS ET AL.

in order to minimize their adverse effect on the environment. The pollution ofsurface and ground waters by acid mine drainage cause serious damage to ecosys-tems. The occurrence of acid waters is related to the solution of sulfide compounds,such as sulfides, in water. Such compounds are very common in volcanic rocksand are originated as metallic deposits precipitated from hydrothermal fluids. Goldis usually found along with these minerals, in exploitable amounts, in relation toadvanced argilic alteration in the andesitic rocks with silica and chalcedony bod-ies. Fluid inclusion data suggests an epithermal system for the ore forming-fluid(Lattanzi et al., 1991; Ruggieri, 1993). Usually, the utilization of the aforemen-tioned compounds creates metal-reach metallurgical wastes, which causes seriousenvironmental problems when they are leached by rain or surface waters. One ofthe most characteristic examples is the case of Sardinia, where after some decadesof production of Pb, Zn, Ag, Cu, Sb, and Ba minerals, enormous volumes of met-allurgical wastes caused serious problems of quality degradation on surface andground waters (Cidu et al., 1997). Also, in Tuscany, the metallurgical wastes orig-inated from the utilization of metal-reach sulfide compounds of Hg, Sb and Fe,created acid mine drainage that has caused serious environmental problems (Ben-venuti et al., 1997). Similar problems occur in South Korea, where the wastes ofpoly-metallic mines produce W, Mo, Fe, Sn, Cu, Mb, Zn and Au, Ag which areconsidered responsible for introducing heavy metals to the environment (Lee et al.,2001; Lee, 2003). In Greece, the intense utilization of silver-bearing galena duringthe period 3000 B.C. to 1989 in the historical mining area of Lavrion, Attica, causedthe accumulation of vast volumes of metallurgical wastes wich are considered to beresponsible for the degradation of the environment (Marinos and Petrascheck, 1956;Konofagos, 1980; Kontopoulos et al., 1995). Stavrakis et al., (1994) and Kontopou-los et al., 1995 mention that the soils of the pit-area, but also in the surrounding areashow high concentrations of Pb, Zn, Sb, Cu, Hg, Cd, As, Fe and Mn. Chronopoulosand Chronopoulou-Sereli, (1986) measured high concentrations of Pb, Zn, Cu, Feand Mn in vegetable fibres. Alexakis and Kelepertsis (1998), and Stamatis et al.(2001) attribute the high metal concentrations in ground water to the geologicalenvironment.

The presence of sulfide minerals has been postulated in the broader Sapes area(Thrace) – northeastern Greece. These minerals occur in volcanic rocks and hy-drothermal fluids.

This study was set up to characterize the environmental hydrochemistry andheavy metal contamination in the broader Sapes area, on the basis of physico-chemical properties of surface and groundwaters occurring in volcanosedimen-tary formations of this area. The results of the study will help identify rockswhich have the potential to produce acid drainage. This work aims also to pre-vent the groundwaters and streams from further pollution in the future, takinginto account that intense gold mining activities are planned to operate in thestudy area in the near future. Figure 1 shows the location map of the studyarea.

HYDROCHEMISTRY OF WATERS OF VOLCANIC ROCKS 377

Figure 1. Principal geologic units and structural features in the study area. Water sampling pointsand stream systems are also shown. (Modified from the geological map 1:200000 of the Institute ofGeological and Mineral Exploration, Greece).


2. Geology of the Study Area

The study area belongs geotectonically to the Rhodope massif located at the south-ern boundary of Xanthi – Komotini basin which was formed during the tertiary uponthe metamorphic rocks of Rhodope massif and the Perirhodopic unit. Accordingto Jacobshagen (1978) the formation of the tertiary basins in the broader area wasa result of the orogenetic phase of Eocene time. The broader Komotini-area, eastof Komotini city, is made up by the following geotectonic units: a) The pre-alpinebedrock of the Rhodope massif, b) The perirhodopic belt (Makri and Drimou-Meliaunit of Mesozoic era) and c) the tertiary basins related to taphrogenic tectonism ofNorth Aegean (Middle to Upper Eocene – Pliocene), (Mposkos et al., 1988). Prin-cipal geologic units, a geologic cross section, structural features as well as watersampling points and streamline network in the study area are shown in Figure 1.

Rocks of Tertiary age present in the study area are intensely altered and un-conformably overly the metamorphic bedrock (Makri unit). According to Mposkoset al. (1988) Makri unit is part or tectonic remnant of the Peri-Rhodopic belt whichis geotectonically related to the Mesozoic geosyncline of Tethys. Makri unit ismade up by the overlying metavolcanosedimentary series or greenschist series andthe underlying metasedimentary series, which mainly is made up by marbles andcalcitic phyllites.

The geologic structure of the broader area is characterized by the presenceof synclinal and antisynclinal structures of many kilometers long. The directionof their axes ranges from 40 to 50◦NE. Generally, folds are vertical-symmetricalwithout inclination of the axial surface (Mposkos et al., 1988). Subvolcanic andintrusive rocks display linear development following tertiary fault zones. Clasticsediments and volcanic rocks were deposited within the study area during a periodcharacterized by intensive synsedimentary – intrusive activity. Pyroxene andesitesand amphibole – biotite andesites, subvolcanic rocks (of rhyo-dakeitic composi-tion) and plutonic rocks are dominant in the study area. According to Michaelet al., (1989a, 1989b), Arikas (1981), Arikas et al. (1993) and Michael et al.,(1995), pyroclastics and lavas in the Rhodope massif locally display paragenesisoriginated from hydrothermal alteration, usually related genetically to basic metaldeposits.

Neogene formations are mainly made up by argillaceous-siliceous minerals. InThrace, lavas, pyroclastics and igneous intrusions are sufficiently altered by hy-drothermal fluids related to sulfer metalliferous deposits (Sideris, 1975; Arikas,1981; Katirtzoglou, 1986). The petrochemical analysis of volcanic rocks indicatesthat they are mineralogically characterized as calc-alkali rocks, namely as basalticandesites, andesites and dacites (Sideris et al., 1991). According to the same au-thors, the average chemical composition of these volcanics is characterized by SiO2

dominance. Basaltic andesites, andesites and dacites contain 55.17%, 60.14% and65.29% of SiO2 respectively. These rocks contain also Al2O3 in a percentage of18.28%, 17.91% and 16.24% respectively, minor amounts of Fe2O3 in a percentage


of 1.07%, 0.77%, and 0.55% respectively, FeO in a percentage of 7.13%, 5.16% and3.71% respectively and smaller amounts of Mn, Mg, Ca, Na, K, Ti, and Cr oxides.The following distinct alteration zones are observed in Sapes area (Marantos et al.,1993): a) The siliceous zone which is composed primarily of quartz (95%) andminor amounts of oxides, e.g. iron oxides (1%) b) The zone of alunite having asmain constituent alunite and minor constituents iron and manganese oxides, c) thesericitic – argillitic zone dominated argillaceous – siliceous minerals and pyrite.This zone contains trace amounts of iron oxides and hydroxides. d) The kaolinitezone which is mainly composed of kaolinite, coexisting with quartz, iron oxidesand hydroxides constitute the minor minerals of this zone.

3. Hydrogeologic Environment

According to the climatic classification of Koppen the climate symbol of the studyarea is “Csa”, (2 < Tc < 18 ◦C and Tw > 22 ◦C), i.e “Mediterranean or Mesother-mal climate type with dry and warm summers and mild winters” (Lambrakis et al.,1999; Petalas et al., 1999), where Tc and Tw is the minimum and maximum annualtemperature accordingly. According to the same authors the average annual precip-itation of the study area is 700 mm and the real annual average evapotranspirationEtr is 300 mm. The runoff is restricted in the period December–March, while dur-ing the rest of the year it is very low to almost null. Floods are rarely produced byintense rainfall (when daily rainfall exceeds 50 mm). The average annual surplusof rain-water, expressing the difference between the relevant values of precipitation(P) and real evapotranspiration (Etr) is 390 mm (Lambrakis et al., 1999; Petalaset al., 1999).

Aquifers are mainly found in the metamorphic formations (especially calc-schists) of Makri unit, in marbles and in volcanosedimentary formations in Maroniabasin which are made up by pyroclastics, sediments, andesites and limestones.Phreatic or semiconfined aquifers are formed in different depths and extent. Thehydrogeological behavior of several geological formations of the study area is asfollows:

The phyllitic system: Shallow aquifers of very low potential are formed withinthe metamorphic rocks phyllites and calc-schists of Makri unit. Additionally, apotential semiconfined aquifer is developed approximately at sea level within thecarbonate rocks (marbles) of Makri unit. As water wells yields are appreciable(greater than 3600 m3/d) groundwater quality is subjected to excessive TDS con-centrations.

Sedimentary rocks: Southern of Perama, extensive areas are occupied by lime-stones of Eocene age containing a potential aquifer. In areas where the lime-stones are fractured and karstified the well yields range from 2400 to 3120m3/d. The specific capacity of the limestone aquifers ranges from 1 m3/h/m to3.5 m3/h/m.


Volcanics (andesite, dacitic-andesite or tuffs): Their hydrologic properties havenot been defined as extensively as those of other lithologies, probably because theirhydrologic variability has discouraged their investigation. Their hydraulic prop-erties differ greatly, and collectively they constitute a complex, heterogeneous,and anisotropic ground-water system. Their permeability is mainly secondary andis attributed to the tectonic activity, as indicated by their petrographic types inthe area. The occurrence of the aquifer system results from the combined ef-fects of fracture system, topography, and weathering. Weathering modifies bothtransmissivity and storage characteristics of these rocks. Transmissivity (T) of thisaquifer system ranges from 57 to 130 m2/d. The storage coefficient ranges from0.002 to 0.0009. The specific capacity ranges from 0.2 to 4.2 m3/h/m (Petalasand Diamantis, 1998). Based on aquifer tests, wells produce as much as 2160m3/d. However, the exploitation potential of the volcanic aquifer system variesfrom place to place. Recently, wells completed in volcanics (andesites) in the areaNW of Perama, with several meters of drawdown, yield 1920 to 2400 m3/d (Peta-las and Diamantis, 1998). These wells penetrated mainly andesites characterizedby recent alteration which developed favourable conditions for the formation ofa potential aquifer system. The natural recharge of the aquifers in all lithologictypes derives entirely from precipitation (including snowmelt) and percolation fromstreams. More rarely recharge may also derive (in phyllite-aquifers) by lateral in-flow from marbles at the contact zones. In the broader area two distinct hydroge-ological units are formed. Water-course aquifers of limited extent are also formedin the study area, within coarse-grained alluvial deposits traversed by perennialstreams.

4. Methods

4.1. GROUNDWATER SAMPLING AND CHEMICAL ANALYSIS

A total of 38 representative samples from surface water and groundwater werecollected for chemical analysis from the study area. The sampling took place duringthe autumn period of 2000, in a network from dug-wells (DW), deep wells (W) andsurface waters (SW). The methods of sample collection and preservation used arespecified in Title 40, Code of Federal Regulations, Part 136 (40 CFR 136) by the U.S.Environmental Protection Agency. Alkalinity, conductivity (EC), temperature, andpH, were measured in the field in order to acquire representative values of ambientaquifer conditions. The samples were collected and stored in each of two 500 mlplastic polyethylene bottles and submitted promptly upon return from the field tothe Laboratory of Hydrogeology of the University of Patras. Chemical analysis wasperformed in one of the two sample bottles, by using a spectrophotometer (modelDR/4000 of the HACK), and the parameters SO4

2−, NO3−, NO2, NH4, PO4

3−,SiO2 were determined. For the determination of Cl− content, titration techniques


were applied by using AgNO3 0,1N (Holl, 1979). One of the two sample bottleswas filtrated prior to the bottling. For sampling purposes, analytical ultra pure acid(dense HNO3) was added to the samples to fix the heavy metals in solution. Majorcations as well as the heavy metals Fe, Mn, Cu, Ni and Zn were measured by usingthe atomic absorption technique (GBC Company – model 908AA). The chemicalanalyses were carried out in the laboratory and the analytical error does not exceed± 5%. PHREEQC software (Parkhurst and Appelo, 1999), was used for calculatingsaturation indices.

4.2. STATISTICAL ANALYSIS

The R-mode multivariate technique is applied in this report on the data of chemicalanalyses to process hydrochemical data and to assist the understanding hydrochem-ical processes and relationships between the several variables (Hakli, 1970). Theprimary aim of factor analysis is the condensation of the maximum information ofthe hydrogeochemical parameters expressed by means of 17 chemical variables,(Table I) in a small number of factors.

The aforementioned technique is routinely used (Voudouris et al, 1997; Stamatis,1999) and it includes the following steps: The first step consists of standardization(mean, xm = 0 and standard deviation σ = 1) of the raw data. This standardizationis given by the formula: zi = (xi−xm)/σ , where zi is the ith value of the standardizedvariable z, xi is the concentration value of variable i . The second step involves thecalculation of correlation coefficients using the formula:

r =Si(xi − xm)(yi − ym)

{[Si(xi − xm)2

] [Si(yi − ym)2

]}1/2

where x and y are the ith values of the standardised variables x and y; xm , ym aretheir respective means.

The correlation coefficients (r) are presented in a matrix form, referred to asthe correlation coefficient matrix. In a third step the determination of eigenvectorsand eigenvalues follow by solving the correlation matrix. By this way, variableswere converted to factors. The isolation of the factor number was based on theKaiser criterion (Kaiser and Cemi, 1979) according to which, no-one of the fac-tors (eigenvectors) has a value (eigenvalue) smaller than 1. In the next step thetable with the eigenvalues was rotated according to the varimax rotation criterion(Davis, 1987) allowing thus better display of the results. Five factors resulted inthis way. Finally, a weighting is assigned to each factor indicating its contribu-tion to each sample (factor scores) necessary for its geographical distribution. We


TAB

LE

I

Bas

icst

atis

tics

ofch

emic

alpa

ram

eter

s(m

g/l)

ofw

ater

sam

ples

Dug

wel

ls(D

W)

Surf

ace

Wat

er(S

W)

Dee

pw

ells

(W)

Var

iabl

eM

inM

axA

vera

geSt

.Dev

.M

inM

axA

vera

geSt

.Dev

.M

inM

axA

vera

geSt

.Dev

.

PH4.

557.

445.

577

5.07

23.

558.

054.

605

4.14

25.

98.

556.

476

6.33

7C

ond

785

5900

2.46

0.90

91.

767.

177

430

3000

1362

.574

6.14

336

074

102.

359.

818

2.58

1.37

2N

a25

530

156.

545

153.

904

3222

574

.548

.832

1912

6029

9.68

247

6.39

7K

170

10.6

6419

.871

1.3

185.

719

4.00

71.

563

21.8

2722

.069

Ca

5142

020

6.09

113

5.38

251

445

167.

688

110.

311

7.5

430

167.

136

155.

555

Mg

20.5

190

82.2

0957

.895

1596

40.9

0625

.177

483

44.2

7334

.708

Cl

2713

5037

2.09

144

6.45

938

572

123.

281

128.

596

22.5

2180

466.

136

846.

157

HC

O3

557

026

5.72

716

5.42

20

388

195.

094

146.

488

3436

119

811

1.61

5SO

452

985

341.

091

302.

883

2313

6039

4.31

338

5.04

725

1324

448.

727

491.

484

NO

30.

558

214

7.36

419

8.95

10.

312

521

.879

36.9

110

356.

409

10.1

48N

H4

00.

60.

152

0.21

70

1.8

0.26

90.

421

00.

370.

133

0.11

8N

O2

00.

150.

039

0.04

30

0.18

0.03

40.

056

00.

010.

002

0.00

4Fe

0.00

90.

060.

036

0.01

60.

012.

80.

332

0.76

80.

006

23.5

4.62

19.

301

Mn

0.00

52.

50.

463

0.94

90.

004

152.

791

5.14

10.

003

4.6

0.97

41.

831

Ni

0.00

40.

085

0.02

50.

022

0.00

60.

150.

034

0.03

90.

006

0.05

0.01

90.

016

Cu

0.00

80.

065

0.03

70.

017

0.00

40.

710.

064

0.17

90.

006

0.03

20.

020.

010

Zn

0.01

60.

290.

066

0.08

80.

005

1.76

0.2

0.46

90.

005

5.6

0.97

21.

761


are thus able to determine the most important elements which affect the observedvariation.

5. Results and Discussion

5.1. GROUNDWATER QUALITY

In Table I, the concentrations of both elements and compounds are given for all of thethree sample categories (DW, SW, and W). Apart form the nitrogen compounds,the rest of the elements and compounds do not present serious differentiationsranging within similar limits. Agricultural practices, including the use of fertilizerscontaining nitrogen, have been recognized in the study area as a key contributorof nitrate. For this reason, nitrate content in ground water showed a great rangeof variability. The most contaminated samples are those from dug wells (DW) andtheir average nitrate content much exceeded EEC/80/778 (1980) and WHO (1984)safe drinking water standard of 50 mg/L. Nitrate contamination of the ground waterfrom deep wells (W) is lower because of the presence of confining units or a thickervadose zone. Other elements and compounds also display high contents, so thatthe environmental and health impacts to be of considerable significance in mostcases. High chloride contents (1000–2200 mg/l) were measured in samples D1,D7, W5 and W6. High chloride contents of samples D1 and D7 may be related tothe presence of evaporitic salts (see also paragraph 5.5), while the high chloridecontents of samples W5 and W7 may be attributed to entrapped relic of an oldseawater intrusion in unflushed parts of aquifers composed of karstified limestonesof Eocene age. Evaporated salts are associated in the study area with sedimentarydeposits of Neogene age (Petalas, 1997). Fossil seawater could have originatedfrom past invasions of the coastal aquifers accompanying rises in sea levels. Theevidences for ancient seawater was primarily from tritium dating (Petalas, 1997).Many samples display low pH values. pH values ranges from 3.5 to 8.5. The higherpH values characterize samples from deep wells (W). The low pH increases thesolubility of heavy metals in the samples D4, D5, D8, S4, S5, S6, S14, S15, W3, W4,W5, W6, W7, W8, W9 where their contents exceed the maximum allowable levels.

5.2. WATER CLASSIFICATION

As it appears in Piper diagram (Figure 2), the three sample categories (DW, SW,W) present similar but complex water types. The order of cation dominance isgenerally Ca2+ > Mg2+, although sodium (Na2+) is the dominant cation in fivewater samples. A small majority of the water samples present the following order ofanion dominance is HCO3

− > SO42− > Cl−. Certain characteristic chemical types

are the following: Ca-Mg-HCO3-SO4, Ca-Mg-SO4-HCO3. Ca-SO4, Ca-Mg-SO4.Ca-Na-Cl-HCO3, Na-Cl.


Figure 2. Piper diagram showing the variation in groundwater and surface water chemistry of thestudy area.

5.3. AQUEOUS CHEMISTRY

As already mentioned, the most pronounced property of waters of the study area istheir acidic character.

The binary diagram of Figure 3a illustrates the relation between pH and SO42−.

Higher sulfate contents of waters are related to low pH. Figure 3b shows that TDS,expressed in the form of EC (Cond, µS/cm), is related also to pH. Moreover, it ap-pears that this relation influences the salt content of surface waters more intenselythan that of groundwaters. The relations between various elements and compoundsreveal that water chemistry is the result of water-rock interaction, while processesas seawater dilution or evaporative concentration play a minor role. Figure 4a showsCl− versus Ca2+ for waters on the study area. The dotted line represents the marinecomposition line relating two end member waters, fresh and sea water. A first obser-vation is that most samples are plotted below freshwater/seawater mixing line, andcalcium is dominant compared to chloride. Figure 4b shows that samples from allwater categories tend to be scattered both above and below the freshwater/seawatermixing line particularly at high concentrations. In some samples sodium contentappears to be in excess compared to chloride, while in some others the opposite isapplied. Figure 4c shows that sulfate ions are dominant compared to chloride ions


Figure 3. Diagrams shown relationships of SO42− and specific electrical conductivity (Cond) versus

pH.

and most samples are plotted below the freshwater/seawater mixing line. As shownin Figure 4d, calcium ions are in excess compared to sulfate ions. Moreover thisfigure shows that the majority of water samples are ordered in the bargain of thesecond dashed line, which connects points with Ca/SO4 ratio, 1:1, implying thatsulfate ions are not originated by gypsum dissolution. The abundance of sulfates(high values) could be attributed to the dissolution of the minerals pyrite (FeS2) andalunite (KAl3(SO4)2(OH)6. Alunite is a very abundant mineral in the study area.However, certain samples as W7 and W11 display Ca/SO4 ratio 1:1. In all caseswhat is observed is a scatter of samples in relation to fresh water/sea water mixingline. In that way, the aspect that the present chemical composition of waters resultsfrom water/rock interaction is confirmed.

As it has already been mentioned, the higher metal concentrations on waters ofthe study area are related to samples with low pH. Acid waters originate by meansof oxidation of sulfide minerals according to the following reaction for pyrite (Hem,1985):

FeS2 + 7 12 O2 + 3 1

2 H2O → Fe(OH)3 + 2H2SO4

or in the general form by the reaction (Cidu et al., 1997):

(Fe, Me)S2 + 3 12 O2 + H2O → Fe2+

aq , Me2+aq + 2SO2−

4 + 2H+,

2Fe2+aq + 1

2 O2 + 2H+ → 2Fe3+aq + H2O.

The oxidation of Fe2+aq to Fe3+

aq , and the instability of Fe3+aq species in solution

when pH rises, leads to precipitation of iron oxyhydroxides, with the consequent


Figure 4. Logarithmic plots of relationships between concentrations of chemical species: Ca2+, Na+,and SO4

2− versus Cl−, and Ca2+ versus SO42−. The dotted line represents the marine composition line

relating two end member waters (the concentration/dilution characteristic for seawater and freshwater).The second dashed line of (d) relates points with a ratio of SO4

2−/Ca2+ = 1:1.

cooprecipitation and/or adsorption of other metals (Johnson, 1986; Kooner, 1993).Additional amounts of SO4

2− in a non-acid environment, contributes to the in-crease of TDS by dissolution of soluble sulfates and this process does not affectpH. Chemical and Biological processes can contribute to the release of metals inwater solutions (Stone and Morgan, 1987). Attenuation of high acid and metalloads may occur through buffering reactions with reactive solid phases in the rock.The principal acid-neutralization mechanisms are represented by dissolution ofcarbonates, hydroxides, and aluminosilicates.

H2SO4 + CaCO3 + 2H2O → CaSO4 · 2H2O + H2O + CO2 (Hem, 1985)

Al(OH)3 gibbsite ↔ Al3+ + 3OH− (Appelo and Postma, 1993)


The previous reaction results in increases in pH, however, it is possible to contributeto increased concentrations of calcium when Al displaces Ca on the exchange sites(exchanger X) according to the reaction:

13 Al3+ + 1

2 Ca − X2 ↔ 13 Al-X3 + 1

2 Ca2+

Thus, the adsorption of aluminum facilitates the continuous dissolution of gibbsiteand results in the creation of buffered conditions. Preceding reactions may occur inthe study area. At present, the chemical evolution of acid waters is difficult to be es-tablished on a quantitative basis, because many of the interrelated factors involved(e.g., the mineralogy and relative abundance of the pH-buffering solids, the sequen-tial nature of pH-buffering reactions, the reaction rates under conditions of dynamicflow) have not been comprehensively evaluated yet (Blowes and Ptacek, 1994).

Increases in alkalinity and pH are observed in certain water samples, as thosefrom wells W1 and W11 that are characterized by the presence of low calciumconcentrations. Two mechanisms may be responsible for the removal of Ca and thealkalinity increase; Dissolution of calcite accompanied by ion exchange processes,or/and hydrolysis of plagioclase (Albite, Anorthite and Analcite) which is controlledby the consumption of CO2. According to the first mechanism, ion exchange processis described by the following equations (Andrews et al., 1994):

Nα2X + Ca2+ → CaX + 2Na+

CaCO3 + H2O → Ca2+ + HCO−3 + OH−

where X− indicates the exchangerThe latter presupposes the presence of calcite and contributes to calcium release

with sodium. Then, the released calcium in solution displaces Na on the exchangesites resulting in rich in sodium solutions unsaturated in respect to calcite. Ac-cording to the second mechanism, increased alkalinity may be caused by meansof hydrolysis of plagioclase feldspars (Albite, Anorthite and Analcite). This mech-anism is also controlled by the consumption of CO2, according to the followingreactions (Andrews et al., 1994):

NaAl2Si3O8 + 5 12 H2O + CO2 → Na+ + HCO−

3 + 12 Al2(Si2O5)(OH)4

+ 2H4SiO4

CaAl2Si2O8 + 3H2O + 2CO2 → Ca2+ + 2HCO−3 + Al2(Si2O5)(OH)4

NaAlSi2O6 + 31/2H2O + CO2 → Na+ + HCO−3 + 1/2Al2(Si2O5)(OH)4

+ H4SiO4

The occurrence of CO2 has been postulated in the basins of tertiary age inMacedonia-Thrace region and is related to sodium rich thermomineral waters in


this region (Lambrakis and Kallergis, 2005). Thus, it is likely that the previousprocesses take place locally in the study area (e.g. locations of wells W2, W11,and D4), and could be indicative of the occurrence of thermomineral waters. Theprevailing of sodium ions in certain samples has also been mentioned in previ-ous paragraphs. In saturated conditions with respect to calcite, the release of Cafrom anorthite dissolution in the solution, as expected from the previous processfor pH higher than 7, results in calcite precipitation and removal from the solutionaccording to the following chemical equation:

Ca2+ + HCO3 → CaCO3 + H+

The formation of the previous minerals results also in the removal of sepiolite(Mg2Si3O75Ohx3H2O) by water.

5.4. WATER–ROCK INTERACTION AND SATURATION INDEX

Figure 5 shows that a considerable number of samples are saturated in respectto minerals calcite and dolomite. Because these samples are distributed in almostthe total of the geological formations, it is evident that calcite and dolomite are

Figure 5. Histograms showing saturations indices of several minerals.


common minerals found in all of them. The higher values of saturation indices(SI > 1) are calculated in water samples from carbonate rocks. All the samples areunsaturated in respect to siderite, anhydrite and gypsum. The diagram of gypsum isnot shown in Figure 5. Finally, all the samples are saturated in respect to mineralsquartz and chalcedony. This is reasonable and expected because SiO2 predominates,particularly in pure volcanic rocks.

5.5. STATISTICAL ANALYSIS AND AREAL DISTRIBUTION OF THE

CHEMICAL PARAMETERS

By using factor analysis a model of five factors is resulted. This model allows thereduction of 17 chemical parameters in five factors that express about 84% of thestatistical information. The high communality values show that the model explainssufficiently the variance of almost all the statistical parameters (Table II).

Factor 1, presents high loadings for the elements Na, K, Cl and for electri-cal conductivity (E.C.). These three elements are generally considered to be ofmarine origin, while the total salinity is expressed by the specific electrical con-ductivity (EC) of the water samples. This factor could be considered as the factorof salinity. The origin of salinity in places located far from the coast could beattributed to cosedimentary entrapment of salts. The second factor is a bipolar

TABLE II

Varimax rotated factor loadings and communalities for Sapes aquifer dataset

Variable Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Communality

pH 0.039 0.883 0.128 −0.028 0.108 0.810EC (µS/cm) 0.932 −0.255 −0.025 −0.105 0.198 0.985Na (mg/L) 0.974 0.013 0.005 −0.01 −0.07 0.954K (mg/L) 0.72 0.138 −0.37 0.009 0.16 0.701Ca (mg/L) 0.313 −0.666 −0.378 −0.114 0.489 0.938Mg (mg/L) 0.518 −0.306 −0.056 0.021 0.558 0.678Cl (mg/L) 0.981 −0.014 0.063 −0.043 0.036 0.969HCO3 (mg/L) 0.257 0.594 0.113 0.241 0.576 0.821SO4 (mg/L) 0.132 −0.792 −0.51 −0.055 0.094 0.916NO3 (mg/L) 0.004 0.111 0.128 −0.092 0.885 0.821NH4 (mg/L) −0.079 −0.185 −0.122 0.92 −0.093 0.91SIO2 (mg/L) 0.111 −0.19 −0.819 0.092 −0.166 0.754Cu (mg/L) −0.035 −0.009 −0.005 0.954 0.06 0.915Fe (mg/L) −0.075 −0.232 −0.829 −0.124 0.103 0.773Mn (mg/L) −0.129 −0.734 −0.194 0.533 −0.045 0.88Zn (mg/L) 0.09 0.019 −0.621 0.354 −0.114 0.532Ni (mg/L) 0.239 −0.842 0.108 0.309 −0.032 0.875% Var 22.7 22.2 14.1 14.0 10.7 83.7


factor and presents high positive loadings for pH and high negative loadings forsulfates and the elements Mn and Ni. This inverse relationship between pH andSO4-Mn-Ni is well correlated to the increased solubility of these elements inlow pH values. The relationship between pH and the REDOX environments isalso well known and accordingly factor 2 expresses the trace elements concentra-tions related to pH, but also the environmental dimension. However, this factorexpresses also the narrow relations between environment, water, air and rocks. pHis accordingly an important control factor for the evolution of the water chemicalcomposition.

Factors 1 and 2 display similar variations related to the total variation (0.22each of them) and accordingly they appear to have similar level of significancein the evolution of water quality. The next two factors display similar variationsand accordingly have analogical level of significance. The rest of the statisticalinformation on metals is attributed to these factors. Factor 3 displays variance 14%,presents high loadings for the elements Fe and Zn which relates with SiO2, whilefactor 4 presents high loadings for Cu and ammonia. Factor 5 with a variance ofabout 10% presents high loadings for nitrate ions. This factor is exclusively relatedto anthropogenic interventions (e.g. intense fertilizations) to the environment andaccordingly to the configuration of water quality. From the precedents it becomesobvious that the presence of metallic minerals in the rocks of the study area is thedecisive factor for the configuration of pH and chemical composition of waters. Thegeographical distribution of factors 1 and 2 is shown in Figure 6. The geographicdistribution of the salinity factor indicates that higher concentrations of salts followa main axis of prevalence with a NW-SE direction, south of Sapes municipality.Also, the increased values of salinity in a distance of a few kilometers from thecoast could be attributed to sea water intrusion by means of natural mechanisms ofsalinization (e.g. mean sea level fluctuation in Pleistocene time). The geographicdistribution of factor 2 is characterized by the presence of two axes. The first axishas a NW-SE direction following the direction of factor 1. The second axis, at first,is almost normal to the first axis, then, it gets a NS direction, south of Perama, wherewater samples display increased metal concentrations. Waters in this area presentalso significantly increased acidity due to the increased sulfides concentrations inthis area.

6. Conclusions

Extended aquifers are formed in volcanosedimentary rocks of the study area. Thetransmissivity (T) of these aquifers is 57 to 130 m2/d, whereas their specific capacityranges from 0.2 to 4.2 m3/h/m. Sulfide compounds of iron, manganese, zinc, nickeland cobalt are very common in the volcanosedimentary rocks. Surface and groundwaters are strongly metalliferous and their hydrochemical facies present similar butcomplex water types as follows: Ca-Mg-HCO3-SO4, Ca-Mg-SO4-HCO3. Ca-SO4,


Figure 6. Areal distribution of the first factor (salinization factor, a) and the second factor (b) in thestudy area.


Ca-Mg-SO4. Ca-Na-Cl-HCO3, Na-Cl. Calcium and magnesium are the dominantcations. Bicarbonates and sulfate ions are the dominant anions. The general decreas-ing order of absolute abundances of major ions in the surface and groundwaters areas follows: HCO3

− > SO42− > Cl− and Ca2+ > Mg2+, although sodium (Na2+)

may be locally the dominant cation. High metal concentrations are related to waterwith low pH. The low pH increases the solubility of heavy metals and their con-tents very often exceed the maximum allowable levels. Sulfide minerals, duringhydrolysis, offer great quantities of minerals in waters while they also lower thepH and therefore they cause acid drainage and consequently quality degradation insurface and groundwaters. The quality of the waters is mainly due to water–rockinteractions. The presence of carbonates creates locally balance conditions in theacid character of the waters as they neutralize the low pH values. Anthropogenicactivities also, e.g. intensive fertilizations, influence adversely water quality.Computer simulation indicates that a considerable number of samples are satu-rated in respect to minerals calcite and dolomite. All the samples are unsaturatedin respect to siderite, anhydrite and gypsum and saturated in respect to mineralsquartz and chalcedony. The increased values of salinity in groundwater a distanceof a few kilometers from the coast could be attributed to the relics an old sea waterintrusion. The water quality problems can be successfully handled by the use offactor analysis. It was concluded that 17 chemical parameters which were measuredin the collected samples can be substituted by five factors which successfully rep-resent (83.7% of the statistical information) the hydrochemical processes as well astheir geographic distribution. Intense gold mining activities which are planned tooperate in the study area in the near future must be appropriately managed as theymay result in leaching of many toxic elements to the surrounding environment.

References

Alexakis, D. and Kelepertsis, A.: 1998, ‘The relationship between the chemical composition-qualityof groundwaters and the geological environment in the East Attica area’, Mineral Wealth 109,9–20, Athens. (In Greek-English summary).

Andrews, N., Fontes, J.C., Aranyossy, J.F., Dodo, A., Edmunds, M.W., Joseph, A. and Travi, Y.: 1994,‘The evolution of alkaline groundwaters in the continental intercalaire aquifer of the Irhazer Plain,Niger’, Water Resources Research 30(1), 45–61.

Arikas, K.: 1981, ‘Subvulcanisch-hydrothermale Mo-Cu-Zn-Pb-vererzugen, S.E. Rhodopen, Nord-griechenland: Petrographie und Geochemie’ TMPM Tschermaks Min. Mitt. 28, 189–205.

Arikas, K., Romstedt, J. and Zahos, S.: 1993, ‘Pegmatites <<metasomatites>> in the metabasiticseries of the east Rhodope massif: Mineralogy-Petrology-Geochemistry’, Geological Society ofGreece Bulletin, Vol. XXVIII(2), 661–673 (In Greek).

Appelo, C.A.J. and Postma D.: 1993, Geochemistry, groundwater and pollution. Rotterdam: A.A.Balkema, 536 pp.

Benvenuti, M., Mascaro, I., Corsini, F., Lattanzi, P., Parrini, P. and Tanelli, G.: 1997, ‘Mine wastedumps and heavy metal pollution in abandoned mining district of Boccheggiano Southern Tuscany,Italy’, Environmental Geology 30(3/4), 238–243.


Blowes, D.W. and Ptacek C.J.: 1994, ‘Acid-neutralization mechanisms in inactive mine tailings’ inJ.L. Jambor and D.W. Blowes (eds.), The Environmental Geochemistry of Sulfide Mine Wastes.Mineralogy Association of Canada, Short Course Handbook 22, pp 271–292.

Chronopoulos, J. and Chronopoulou-Sereli, K.: 1986 ‘Schwermetalltoleranz von Crocus sieberi, Aris-arum vulgare und Cyclamen graecum in Lavrion (Attika)’, Verhandlungen der Gesellschaft furOkologie, XIV, 357–360, Gottingen.

Cidu, R., Caboi, R., Fanfani, F. and Frau, F.: 1997, ‘Acid drainage from sulfides hosting gold miner-alization (Furtei, Sardinia)’ Environmental Geology 30(3/4), 231–237.

Davis, J.C.: 1986, Statistics and Analysis in Geology, 2nd ed., John Wiley & Sons, New York, 656pp.

E.E.C.: 1980, ‘Council directive (80/778) on the quality of drinking water’.Holl, K.: 1979, Wasser: Untersuchung, Beurteilung, Aufbereitung, Chemie, Bakteriologie, Virologie,

Biologie, pp. 515, Berlin-New York (Walter de Gruyter).Hakli, T.A.: 1970, ‘Factor analysis of the sulphide phase in mafic/ultramafic rocks in Finland’, Bull.

Geol. Soc. Finland 42, 109–118.Hem, D.: 1985, ‘Study and Interpretation of the chemical characteristics of natural water’, U.S. Geol.

Survey Water Supply Paper 2254, 3rd ed.Jacobshagen, V.: 1978, ‘Structure and geotechnic evolution of the Hellenides’, in G. Kallergis (ed.),

Proceedings of the VI Colloqium on the geology of the Agean region, Athens, Greece, 1977, Vol.III, 1355–1367.

Johnson, C.A.: 1986, ‘The regulation of trace element concentra-tions in river and estuarine waterscontaminated with Acid mine drainage: The adsorption of Cu and Zn on amorphous Fe oxyhy-droxides’, Geochim. Cosmochim. Acta 50, 2433–2438.

Kaiser, H.F. and Cerny, B.E.: 1979, ‘Factor analysis of the image correlation matrix’, Educationaland Psychological Measurement, 711–714.

Katirtzoglou, K.: 1986, ‘Metallogeny of the tertiary sulfide occurrence in Esimi area, Evros No-marchy’ Ph.D. Thesis, Department of Geology, University of Athens, Greece, 176 pp. Athens[in Greek].

Konofagos, E.C.: 1980, ‘Le Laurium antique et la technique Grecque de la production de l’argent’.National Technical University, Athens, 458 pp. (in French).

Kontopoulos, A., Komnitsas, K., Xenidis, A. and Papassiopi, N.: 1995, ‘Environmental characteriza-tion of the sulphidic tailings in Lavrion’, Minerals Engineering 8(10), pp. 1209–1219.

Kooner, Z.S.: 1993, ‘Comparative study of adsorption behaviour of copper, lead, and zinc onto goethitein aqueous systems’, Environmental Geology 21, 242–250.

Lambrakis, N., Petalas, C., Diamantis, J. and Kallergis, G.: 1999, ‘Hydrogeological conditions andwater quality of the volcanic aquifers of the broad area of Sapes, Thrace’, Proceedings of the 5thPanhellenic Hydrogeological Conference, Cyprus Association of Geologists and Mining Engi-neers Bulletin 9, 205–216, Nicosia (in Greek).

Lamprakis, N. and Kallergis, G.: 2005, Contribution to the study of Greek thermal springs: Hydro-geological characteristics and origin of thermal waters’, Hydrogeology Journal (in press).

Lattanzi, P., Ruggieri, G. and Tanelli, G.: 1991, ‘Preliminary fluid inclusion data on epithermal goldmineralization at Serrenti-Furtei, SW Sardinia, Italy’, Plinius 5, 127–128.

Lee, C.H.: 2003, ‘Assessment of contamination load on water, soil and sediment affected by theKongjujeil mine drainage, Republic of Korea’, Environmental Geology 44, 501–515.

Lee, C.H., Lee, H.K. and Lee, J.C.: 2001, ‘Hydrogeochemistry of mine, surface and groundwatersfrom the Sanggok mine creek in the upper Chungju Lake, Republic of Korea’, EnvironmentalGeology 40, 482–494.

Marantos, I., Michael. K. and Perdikatsis, B.: 1993, ‘Preliminary study for industrial minerals inaltered vulcanites of Sapes area, Rhodope Nomarchy, Greece’, Geological Society of GreeceBulletin, Vol. XXVIII(2), 317–328, Athens (in Greek).


Marinos, G. and Petrascheck, W.E.: 1956, ‘Lavrium’, Geol. and Geoph. Research 4(1). Inst. Geol.and Subsurface Research, Athens.

Michael, C., Constantinides, D., Ashworth, K., Perdikatsis, V. and Demetriades, A.: 1989a, ‘The Kirkivein polymetallic mineralization, NE Greece’, Geol. Rhodopica 1, 366–373.

Michael, C., Katirtzoglou, C., Perdikatsis, V. and Constantinides, D.: 1989b, ‘The polymetalic min-eralization of the Pefka area, Evros Nomarchy, Greece’, Geol. Rhodopica 1, 322–329.

Michael, C., Perdikatsis, V., Dimou, E. and Marantos, I.: 1995, ‘Hydrothermal alteration and ore de-position in epithermal precious metal deposit of Agios Demetrios, Konos area, Northern Greece’,Geol. Soc. Greece, Sp. Publ. 4, 778–780, Athens.

Mposkos, E., Papadopoulos, P. and Perdikatsis B.: 1988, ‘The crystalline basement of the Rhodopicmassif east of Komotini’, Geological Society of Greece Bulletin, Vol. XX(2), 259–273, Athens(in Greek).

Petalas, C.: 1997, ‘Analysis of aquifer systems in the heterogeneous coastal plain of Rhodope region’,Doctoral Dissertation, University of Thrace, Department of Civil Engineering, Xanthi, Greece (inGreek).

Petalas, C. and Diamantis, I.: 1998, ‘Chemical characteristics of groundwaters in hard rocks ofAlexandroupoli area. Geological Society of Greece Bulletin, Vol. XXXII (4), 59–68, Athens (inGreek).

Petalas, C., Diamantis, J., Lambrakis, N. and Kallergis, G.: 1999, ‘Interaction of natural waters withthe geological formations of Makri unit and the graben of Maronia –Thrace’, Proceedings ofthe 5th Panhellenic Hydrogeological Conference, Cyprus Association of Geologists and MiningEngineers Bulletin 9, 423–438, Nicosia (in Greek).

Parkhurst, D.L. and Appelo, C.A.J.: 1999, ‘User’s guide to PHREEQC (Version 2) – A computerprogram for speciation, batch-reaction, one-dimensional transport, and inverse geochemical cal-culations’, U.S. Geological Survey Water-Resources Investigations Report 99–4259, 310 pp.

Ruggieri, G.: 1993, ‘Ore genesis in epithermal environment: The examples of Furtei Sardinia), Frassineand La Campigliola (southern Tuscany) gold mineralization’, Plinius 9, 134–139.

Sideris, K., Katirtzoglou, K. and Hatzikiriakou, A.: 1991, ‘Petrology and geochemistry of the tertiaryvolcanic rocks of Esimi-Kirki area, East Rhodope’, Geological Society of Greece Bulletin Vol.XXV(2), 19–31, Athens (in Greek).

Sideris, K.: 1975, ‘Browse of especial characters in plutonic and volcanic rocks of east Rhodopiczone – relationships between vulcanicity and plutonic activity’, Post doctorate dissertation, De-partment of Geology, University of Athens, 108 pp.

Stamatis, G., Voudouris, K. and Karefalakis, F.: 2001, ‘Groundwater pollution by heavy metals inhistorical mining area of Lavrio, Greece’, Water, Air, and Soil Pollution 128, 61–83.

Stavrakis, P., Vergou-Vichou, K., Fosse, G., Makropoulos, V., Dimitriadis, A. and Vlachogiannis, N.:1994, ‘A Multidisciplinary Study on the Effects of Environmental Contamination on the HumanPopulation of the Lavrio urban area; Hellas’, Proceedings of the 6th International Conference onEnvironmental Contamination, Delphi, Greece, October 1994, 1–20.

Stone, A.T. and Morgan, J.J.: 1987, ‘Reductive dissolution of metal oxides’, in W. Stumm (ed.),Aquatic Surface Chemistry, New York, John Wiley & Sons, 221–254.

Voudouris, K., Lambrakis, N., Papatheodorou, G. and Daskalaki, P.: 1997, ‘Factor analysis for re-vealing hydrochemical parameters in Pleistocene aquifers of NW Achaia, Peloponnesos, Greece’,Mathematical Geology 29(1) 43–59.

W.H.O.: 1984, ‘Guidelines for drinking-water quality’, Vol. 2, Health criteria and other supportinginformation. Geneva.

Documents

HYDROCHEMISTRY OF WATERS OF VOLCANIC ROCKS: THE … · 2011-02-24 · HYDROCHEMISTRY OF WATERS OF VOLCANIC ROCKS: THE CASE OF THE VOLCANOSEDIMENTARY ROCKS OF THRACE, GREECE C. PETALAS1,∗,N.LAMBRAKIS2