Reactive Transport Modeling of Thermal Column ... - ACS Publications

Sep 24, 2014 - Aquifer thermal energy storage (ATES) systems are increasingly being used to acclimatize buildings and are often constructed in aquifer...
2 downloads 0 Views 2MB Size
Article pubs.acs.org/est

Reactive Transport Modeling of Thermal Column Experiments to Investigate the Impacts of Aquifer Thermal Energy Storage on Groundwater Quality Matthijs Bonte,*,†,§ Pieter J. Stuyfzand,†,‡ and Boris M. van Breukelen‡ †

KWR Watercycle Research Institute, P.O. Box 1072, 3430BB, Nieuwegein, The Netherlands Critical Zone Hydrology Group, Department of Earth Sciences, VU University Amsterdam, De Boelelaan 1085, 1081HV, Amsterdam, The Netherlands.



S Supporting Information *

ABSTRACT: Aquifer thermal energy storage (ATES) systems are increasingly being used to acclimatize buildings and are often constructed in aquifers used for drinking water supply. This raises the question of potential groundwater quality impact. Here, we use laboratory column experiments to develop and calibrate a reactive transport model (PHREEQC) simulating the thermally induced (5−60 °C) water quality changes in anoxic sandy sediments. Temperature-dependent surface complexation, cation-exchange, and kinetic dissolution of K-feldspar were included in the model. Optimization results combined with an extensive literature survey showed surface complexation of (oxy)anions (As, B, and PO4) is consistently exothermic, whereas surface complexation of cations (Ca and Mg) and cationic heavy metals (Cd, Pb, and Zn) is endothermic. The calibrated model was applied to simulate arsenic mobility in an ATES system using a simple yet powerful mirrored axi-symmetrical grid. Results showed that ATES mobilizes arsenic toward the fringe of the warm water bubble and the center of the cold water bubble. This transient redistribution of arsenic causes its aqueous concentrations in the cold and warm groundwater bubbles to become similar through multiple heating cycles, with a final concentration depending on the average injection temperature of the warm and cold ATES wells.



INTRODUCTION Aquifer Thermal Energy Storage (ATES) systems use the underground to store or harvest renewable energy for heating or cooling purposes. 1 Utilization of ATES is rapidly growing in western European countries, and especially in The Netherlands where the number of licensed ATES systems increased from 29 systems in 1995 to around 1800 systems in 2012. 2 ATES systems typically comprise two wells (called a well doublet), which serve both as extraction and injection well, depending on the season. In summer, groundwater is extracted from the “cold” well and used to cool buildings. This raises the temperature of the groundwater which is subsequently injected in the “warm” well. In winter, the flow direction reverses and groundwater is extracted from the warm well and used to heat the building. This lowers the temperature of the groundwater before it is reinjected into the cold well. The maximum injection temperature in countries like The Netherlands, Germany, and Austria is legally limited to 20−25 °C 3 but most systems work up to 15 °C. Systems with higher operational temperatures (up to 110 °C) are less frequently reported,4,5 but interest in these systems is growing as a technology of storing waste heat for later reuse. ATES systems are often realized in the same aquifers used for drinking water © 2014 American Chemical Society

production, which raises questions about the effects of ATES on water quality.6,7 Published research showed ATES can influence groundwater quality by mixing of different water quality types,8 temperature changes,4,5 or intrusion of oxygen or degassing.9,10 Most of these studies focused on operational aspects, such as scaling in relation to mineral precipitation.4,5,9,11 A number of laboratory studies showed that the microbial community structure and groundwater fauna was not significantly impacted by temperature increases up to 30 °C,12−14 which was confirmed in data from a field ATES site. 14 However, the rates of kinetically restricted biochemical processes do show a clear temperature dependency, as observed in laboratory experiments for sulfatereduction12,15 and oxidation of ferrous iron,16 and in the field for pyrite oxidation. 17 In our previous research, we employed column experiments to investigate the possible effects of thermal changes in anoxic unconsolidated sandy sediments and showed that particularly Received: Revised: Accepted: Published: 12099

May 20, 2014 September 10, 2014 September 24, 2014 September 24, 2014 dx.doi.org/10.1021/es502477m | Environ. Sci. Technol. 2014, 48, 12099−12107

Environmental Science & Technology

Article

arsenic and to a lesser degree boron, silica, potassium, phosphorus, and dissolved organic carbon (DOC) became increasingly mobile upon a temperature increase. 18 Also, sulfate-reduction showed a strong temperature-dependence. 12 The aim of the current work is to simulate the experimental results of the column tests 18 with a hydrogeochemical reactive transport model using the PHREEQC 19 code to gain a quantitative understanding of the temperature dependence of silicate weathering, cation exchange, and surface complexation reactions. We extended the Dzombak and Morel 20 (D&M) database, available in PHREEQC for modeling surface complexation, to include reaction enthalpies for a selection of relevant solutes, which to our knowledge has not been attempted or published before. The calibrated model was subsequently applied to simulate arsenic mobility in a representative ATES system at three injection temperatures using a novel computationally efficient modeling concept composed of two mirrored axi-symmetrical flow tubes. This setup allows for gaining a fast and numerically efficient understanding of the prevailing hydrochemical processes in an ATES system without the need of full 3-D hydrogeochemical transport modeling.

Table 1. Initial Pore Water, Influent, and Sediment Compositions Used in the PHREEQC Simulationsa



EXPERIMENTAL METHODS The experimental methods and data, and conceptual interpretation, used to construct the reactive transport model is described in detail in Bonte et al.18 A comprehensive summary is provided here. Sediments and water used in the column experiments were collected at Scherpenzeel (Netherlands) from a depth of 34−36 m-SL. The sediment was collected from an unconsolidated sandy aquifer (Sterksel formation, Early to Middle Pleistocene 21). The sample consisted mainly of quartz sand (>90%), with minor fractions of K-feldspar, clay minerals, organic matter, carbonates, pyrite, and reactive iron (Table 1 and Supporting Information (SI) Table S1). Following sample collection and homogenization in a glovebox under N2 atmosphere, four identical sediment columns (with length = 440 mm and diameter=66 mm) were placed in the experimental setup (SI Figure S1) at temperatures of 5 °C (representing cold storage), 11 °C (ambient temperature), 25 °C (maximum allowed regular ATES), and 60 °C (high temperature ATES). Influent (Table 1) was derived from groundwater collected from the same borehole from which the sediment cores were taken. The columns were flushed and the effluents were frequently sampled. The residence time in the columns was 1 day and a thermodynamic equilibrium was assumed for cation-exchange and surface complexation. Afterward, a tracer test was conducted using NaCl to determine the dispersivity (α) in the cores. This yielded α = 9 cm for the 11 °C column, α = 10 cm for the 5 and 25 °C columns, and α = 24 cm for the 60 °C column (SI Figure S2 and Table S2).



water

initial pore water

influent

pH alkalinity (meq/L) dissolved oxygen NO3 NH4 SO4 P DOC K Si Ca Mg Sr Fe(II)b Mn As(III)b B

7.2 1.8