doi:10.1016/j.chemgeo.2007.10.007
Copyright © 2007 Elsevier B.V. All rights reserved.
Inverse and forward modelling of groundwater circulation in a seismically active area (Monferrato, Piedmont, NW Italy): Insights into stress-induced variations in water chemistry
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
C. Federicoa, 
, 
, L. Pizzinob, D. Cintib, S. De Gregorioa, R. Favaraa, G. Gallib, G. Giudicea, S. Gurrieria, F. Quattrocchib and N. Voltattornib
aIstituto Nazionale di Geofisica e Vulcanologia, sezione di Palermo, via U. La Malfa 153, 90146 Palermo, Italy
bIstituto Nazionale di Geofisica e Vulcanologia, sezione di Roma 1, via di Vigna Murata 605, 00143 Roma, Italy
Received 8 August 2006;
revised 16 October 2007;
accepted 17 October 2007.
Editor: J. Fein.
Available online 26 October 2007.
Abstract
Reaction path modelling, coupled with preparatory inverse modelling, was applied to test this model's ability to reproduce the wide compositional range of ground waters circulating in a restricted area in Piedmont, Italy. This approach is based on the assumption that the chemistry of groundwater evolves through a series of partial equilibria with secondary minerals until it reaches its final composition. PHREEQC [Parkhurst, D.L., Appelo, C.A.J., 1999. User's guide to PHREEQC-A computer program for speciation, reaction-path, 1D-transport, and inverse geochemical calculations. U.S. Geological Survey Water-Resources Investigations Report, pp. 99-4259] and EQ3/6 [Wolery, T.J., Daveler, S.A., 1992. EQ6, A Computer Program for Reaction Path Modeling of Aqueous Geochemical Systems: Theoretical Manual, User's Guide and Related Documentation (version 7.0). Report UCRl-MA-110662 PT IV. Lawrence Livermore National Laboratory, Livermore, California] software packages were used to effect simulations. Reaction-path modelling was performed in time mode, taking into account the different rates of dissolution of each dissolving mineral.
Data from literature regarding the kinetic parameters of dissolving minerals and the mineralogical composition of the host-rock were used. The results of the reaction-path modelling show that the composition of the analysed water samples was adequately reproduced, notwithstanding the hydrogeological complexity of the studied area. Modelling results provided very different water compositions as an effect of the chemical maturity, the physico-chemical parameters (fCO2, fO2, and temperature) and the variable amounts of gypsum among dissolving rock-forming minerals, which occur in Miocene levels of the sedimentary sequence. Further variability is related to the occasional contribution of brackish waters trapped in euxinic marly sediments, locally sealed by overlying clays, that have assumed an artesian character. The composition of some of the water samples can only be predicted by simulation runs performed at a temperature higher than that of the outlet (40 °C). These warm waters probably circulate in a restricted area near the town of Nizza Monferrato. The same area has recently been affected by moderate seismicity, which has been accompanied by changes in either the temperature or chemistry, or both, of the ground waters. The changes recorded, interpreted as having been triggered by variations in the local/regional stress load and/or seismic activity, have to be ascribed to the vertical heterogeneity of the aquifers, where waters of different temperature, salinity and chemical composition circulate and occasionally mix.
Keywords: Reaction-path modelling; EQ3/6; Inverse modelling; Tertiary Piedmont Basin; Monferrato; Seismicity
Fig. 1. Structural sketch map of North-Western Italy (modified from Carrapa and Garcia-Castellanos, 2005). The box indicates the investigated area. LA: Ligurian Alps; TPB: Tertiary Piedmont Basin; AM: Alto Monferrato; M: Monferrato; IL: Insubric line; VVL: Villalvernia-Varzi line; VGT: Val Gorrini thrust; SVZ: Sestri-Voltaggio zone; VG: Voltri Group.
Fig. 2. Lithological sketch map and location map of sampled wells and springs (ARPA Piemonte, 1990). Tracks of the geological sections of Fig. 5 are also indicated. The area where well waters have been warming up since 2000 is indicated by an ellipse.
Fig. 3. δ18O vs. δD scatter plot. The lines representing both the global meteoric water line (MWL δD = 8δ18O + 10; Craig, 1961) and the Eastern Mediterranean water line (EMWL, δD = 8δ18O + 22; Gat and Carmi, 1970) are also plotted.
Fig. 4. a) Triangular diagram representing the relative proportion of Na + K, Mg and Ca in Monferrato waters, where the fields of alkali-dominated, Mg-dominated and Ca-dominated waters are indicated. Relative concentrations are on a weight basis; b) Cl–SO4–HCO3 triangular diagram. The fields regarding chloride waters, bicarbonate waters and sulphate-rich waters have been differentiated.
Fig. 5. a, b Interpretative geological sections in the study area. The tracks of the sections have been drawn in Fig. 2. The vertical scale is magnified. Sampled waters tap different aquifers, where different physico-chemical conditions hold. Artesian wells 12, 22 and 23 and spring 4 tap the locally confined aquifer, hosted in clayey and marly sediments of Oligocene–Miocene age, where CH4 pockets occur associated to brackish waters. Wells 6, 7, 8, 9 and 10 tap the phreatic portion of the aquifer hosted in the same marly sediments, where the impermeable cover of overlying Miocene clays is lacking. These wells have also been contaminated by waters either circulating in quaternary alluvial deposits or in Miocene evaporitic deposits. Wells 1 and 31 tap gypsum-bearing clays and marly clays of Upper Miocene age.
Fig. 6. Na-pH binary diagram. The results of the two reaction path models have been drawn as curves 1 and 2.
Fig. 7. Activity diagram Log (Mg/H+2) vs. Log (SiO2(aq)), showing stability compositional fields of some relevant secondary mineral phases, drawn at 1 bar pressure and both 15 (dotted line) and 25°C (solid line) using thermodynamic data from the data0.com.R2 database (Wolery and Daveler, 1992). For comparison, the saturation lines of amorphous silica at different temperatures have been plotted. The compositional lines (identified by numbers), relative to model solutions computed through reaction path modelling, are also shown. The results of the different reaction path simulations have been drawn; they were performed at the conditions indicated at the bottom of the graphs.
Fig. 8. Computed saturation index for some secondary minerals. Dashed lines represent the equilibrium condition; over saturation with respect to indicated solid phases is above the line and under saturation below.
Fig. 9. The saturation index of some selected minerals has been plotted versus temperature, following the method proposed by Reed and Spycher (1984). Plots a, b, c, d, and e refer to samples 1, 31, 22, 23, and 4 respectively.
Fig. 10. a,b Na–Si–Mg triangular plot. Relative proportions are on a weight basis. The model generally tends to evolve towards the Mg, after the Si and, finally, the Na corners, due to the leaching of both chlorite and smectite dissolution and finally by plagioclase dissolution. The star represents brackish water, 83% meteoric, and 17% marine.
Fig. 11. a,b Si–Ca–SO4 triangular diagrams. Relative concentrations are on a weight basis. The points representing the leaching solution for the model lines 5 and 6 are connate water of marine origin, partly diluted with shallow water, SO4-depleted due to the very reducing conditions holding in organic-C-rich sediments.
Fig. 12. a) Cumulative Log of moles of minerals destroyed during reaction path modelling and b) Cumulative Log of moles of newly-forming secondary phases, plotted versus the progress variable ξ, for the simulation 1.
Fig. 13. Na–Mg binary plot. The results of reaction path modelling have also been plotted (curves 1, 2, 3 and 5). Samples from well 31, collected before the June 2004 earthquake, have been highlighted by an ellipse.
Fig. 14. Na–Si binary plot. The results of reaction path modelling have also been plotted (curves 1, 2 and 3).
Fig. 15. Na–SO4 binary plot. The star indicates the composition of brackish water (94% meteoric, 6% marine).
Fig. 16. Time trends of Cl, SO4 and As contents in sample no. 6.
Fig. 17. Temporal trends of HCO3/SO4 ratios and δ18O in well 31.
 | (683 K)
 |
Table 2. Major, minor and trace element contents and water δ
18O in some selected sites from 2003 to 2005
Table 1.
Major, minor and trace element contents in ground waters analysed from the 2003 preliminary survey
Stable water isotopes have also been listed.
Table 2.
Results of inverse modelling
The amount of dissolving and precipitating minerals is expressed as mol/kg water.
Table 3.
Kinetic parameters used in the EQ6 simulations
Specific surface areas were determined by the BET method, if not otherwise specified. Activation energy values were applied in the Arrhenius equation (
) to compute dissolution rates at different temperatures. In the Arrhenius equation, ki stands for kinetic constant for the ith aqueous species involved in the formation of activated complexes and B is a pre-exponential factor. In the rate expression for gypsum dissolution, cs and ce are Ca concentrations at the surface and at equilibrium, respectively.
Corresponding author. Fax: +39 091 6809449.
Abstract
Purchase PDF (2161 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (0)
Purchase PDF (310 K)
