Geochemical processes conceptual models for reactive transport in soil and groundwater

Môi trường ngày càng ô nhiễm nặng, việc chung tay bảo vệ là việc của tất cả mọi người trên trái đất này. Sau đây Dịch thuật Hồng Linh dịch thuật tiếng anh giá rẻ xin giới thiệu một số thuật ngữ tiếng anh ngành môi trường. > English Việt Nam absorptionabsorbent (sự, quá trình) hấp thụchất hấp thụ absorption field mương hấp thụ xử lý nước từ bể tự hoại acid deposition mưa axit acid rain mưa axit

Deutsche Forschungsgemeinschaft Geochemical Processes Conceptual Models for Reactive Transport in Soil and Groundwater Geochemical Processes: Conceptual Models for Reactive Transport in Soil and Groundwater Research Report Deutsche Forschungsgemeinschaft (DFG) Copyright © 2002 WILEY-VCH Verlag GmbH & Co KGaA, Weinheim ISBN: 978-3-527-27764-3 Deutsche Forschungsgemeinschaft Geochemical Processes Conceptual Models for Reactive Transport in Soil and Groundwater Edited by Horst D Schulz and Georg Teutsch Research Report Deutsche Forschungsgemeinschaft Kennedyallee 40, D-53175 Bonn, Federal Republic of Germany Postal address: D-53170 Bonn Phone: ++49/228/885-1 Telefax: ++49/228/885-2777 E-Mail: postmaster@dfg.de Internet: http://www.dfg.de This book was carefully produced Nevertheless, editors, authors and publisher not warrant the information contained therein to be free of errors Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate Library of Congress Card No.: applied for A catalogue record for this book is available from the British Library Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at http://dnb.ddb.de ISBN 3-527-27764-1 © 2002 WILEY-VCH Verlag GmbH & Co KGaA, Weinheim Printed on acid-free paper All rights reserved (including those of translation into other languages) No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into machine language without written permission from the publishers Registered names, trademarks, etc used in this book, even when not specifically marked as such, are not to be considered unprotected by law Cover Design and Typography: Dieter Hüsken Composition: Hagedorn Kommunikation, Viernheim Printing: betz-druck gmbh, Darmstadt Bookbinding: J Schäffer GmbH & Co KG, Grünstadt Printed in the Federal Republic of Germany Contents Modelling Contaminant Transport in Anthropogenic Soil: Reconstruction of Spatial Heterogeneity by Analysing the Relations of Adjacent Pedofacies Kai Uwe Totsche, Ingrid Kögel-Knabner and Harald Weigand 1.1 Introduction 1.2 Characterising Spatial Heterogeneity of Soils at Anthropogenic Sites: the Testfeld Süd Reconstruction of Spatially Variable Pedofacies with a Structure-imitating Stochastic Approach Based on Markov Processes Markov Chain Theory Continuous-lag Markov Chains 10 10 Assessing the Risk for PAH Deep Seepage at Industrial Contaminated Sites 14 1.5 Summary and Conclusion 17 1.6 References 18 1.3 1.3.1 1.3.2 1.4 V Contents Concepts for Modelling of Heterogeneous Flow Processes in Soil Columns on the Basis of Tomographic Radiotracer Experiments Michael Richter 20 2.1 Introduction 21 2.2 Technology and Applicability of the Positron Emission Tomography (PET) for Transport Studies in Soil Columns 21 2.3.1 2.3.2 Typical Results of PET-Studies of the Hydrodynamics in Soil Columns Experimental PET-Measurements of Tracer Distribution in the Model Soil Column 24 24 25 2.4 Calculation of Velocity Distributions 28 2.5 2.5.1 2.5.2 2.5.3 2.5.4 Concepts of Modelling of Flow Processes in Soil Columns Estimation of Parameters by Inverse Modelling Partition of the Column in Regions with Different Flow Characteristics Modelling with Reference to the Dispersion Model Support of Conventional Measuring Technique 2.3 30 30 31 33 34 2.6 References 38 Upscaling of Hydraulic and Hydrogeochemical Aquifer Parameters Using an Approach Based on Sedimentological Facies Thomas Ptak and Rudolf Liedl 39 3.1 Introduction 39 3.2 3.2.1 3.2.3 The Three-dimensional Reactive Transport Modelling Approach Facies-based Characterization of Hydraulic and Hydrogeochemical Aquifer Properties Generation of Three-dimensional Facies and Fields of Hydraulic and Hydrogeochemical Aquifer Parameters Modelling of Flow and Reactive Transport 41 42 44 47 3.3 Example of Application 49 3.4 Conclusions and Future Work 53 3.5 References 53 3.2.2 VI Contents DiffMod7 – Modelling Oxygen Diffusion and Pyrite Decomposition in the Unsaturated Zone Based on Ground Air Oxygen Distribution Henrik Hecht, Martin Kölling and Norbert Geisler 55 4.1 Introduction 55 4.2 Data Basis for Model Development 4.3 4.3.1 4.3.1.1 4.3.1.2 4.3.1.3 4.3.2 4.3.3 4.3.4 4.3.5 4.3.6 The Model (DiffMod7) Model Concept Diffusive Transport Pyrite Weathering Convective Transport Computer Implementation of Model and User Interface Example Modelling Column Experiments Running Scenarios Field Test 58 58 58 60 61 61 62 64 67 72 4.4 Discussion 74 4.5 Summary 76 4.6 References 77 Speciation and Sorption for Risk Assessment: Modelling and Database Applications Vinzenz Brendler, Thuro Arnold, Sture Nordlinder, Harald Zänker and Gert Bernhard 79 5.1 Introduction 79 5.2 5.2.1 5.2.1.1 5.2.1.2 5.2.1.3 5.2.2 Geochemical Speciation and Sorption The Concept of “Smart Kd” Application Case: Sorption onto Rocks Application Case: Colloids in Mines Application Case: Uranium Migration Mineral-Specific Sorption Database 80 81 83 85 87 91 5.3 References 93 57 VII Contents New Geochemical Simulator Rockflow-RTM: Development and Benchmarking Abderrahamne Habbar, Olaf Kolditz and Werner Zielke 95 6.1 Introduction 96 6.2 6.2.1 6.2.2 Governing Equations Nonequilibrium Equations Equilibrium Equations 96 96 98 6.3 Numerical Method 99 6.4 Software Concept 6.5 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 6.5.6 Examples Nitrification Process in a Porous Column Geochemical Nonequilibrium Effects TCE Transformation Matrix Diffusion Two-Member Decay Chain in Fracture-Matrix System Two-Member Decay Chain in Fracture-Matrix System 6.6 Conclusions 6.7 References 114 Modelling Reactive Transport of Organic Solutes in Groundwater with a Lagrangian Streamtube Approach Michael Finkel, Rudolf Liedl and Georg Teutsch 115 7.1 Introduction 7.2 7.2.1 7.2.2 7.2.3 Reactive Transport Model SMART Streamtube Approach Accounting for Reactive Processes Numerical Evaluation of Breakthrough Curves 117 117 119 120 7.3 7.3.1 7.3.2 7.3.3 7.3.4 7.3.5 7.3.5.1 7.3.5.2 Reactive Transport of Phenanthrene and Terrasurf G50 Conceptual Model of Relevant Processes Remediation Scenario Process Parameters Conservative Transport Description Reactive Transport Simulations Influence of Aquifer Properties on Transport of PHE Impact of Non-Ionic Surfactant TG50 on Transport of PHE 123 123 125 126 127 127 127 130 VIII 100 101 102 104 105 107 110 111 113 115 Contents 7.4 Summary and Conclusions 7.5 References 132 Conception of a GIS-Based Data Model for Combined Hydrochemical and Hydraulic Balance Calculations in Pleistocene Landscapes – an Approach of Regionalization Christoph Merz, Peter Schuhmacher, Jörg Steidl and Andreas Winkler 135 8.1 Introduction 8.2 Material and Methods 136 8.3 8.3.1 8.3.2 8.3.2.1 8.3.2.2 8.3.2.3 8.3.3 Results Stoebber Watershed Oderbruch Introduction Material and Methods Results Ucker Watershed 8.4 Discussion 150 8.5 References 152 The “Virtual Aquifers” – Concept as a Tool for Evaluation of Exploration, Remediation and Monitoring Strategies Dirk Schäfer, Andreas Dahmke, Olaf Kolditz and Georg Teutsch 154 9.1 Introduction 9.2 9.2.1 9.2.3 Examples for the Use of the Virtual Aquifer Concept Effect of Screening and Pumping Rate on Measured Concentrations in a Heterogeneous Aquifer The Effect of Mixing in an Observation Well on Evaluation of Natural Attenuation Processes in a Heterogeneous Aquifer Monitoring of Natural Attenuation in a Heterogeneous Aquifer 9.3 Problems and Requirement for Additional Scientific Research 168 9.4 The “Virtual Aquifer” Project 170 9.5 References 171 9.2.2 131 136 139 139 140 140 142 142 148 154 156 156 160 166 IX Contents 10 Two-Dimensional Two-Step Modelling of 250 Years of Transport and Reactions in a Virtual Anoxic Aquifer (Oderbruch, Eastern Germany) Gudrun Massmann and Horst D Schulz 173 10.1 10.1.1 10.1.2 Introduction 174 Why 2D-Modelling of Transport and Reactions? 174 Why the Study Site Oderbruch? 175 10.2 The Study Site: Hydraulic Conductivity and Iron(III) Concentration of the Aquifer 176 10.3 10.3.1 10.3.2 Principle Structure of the Two-Step Model 181 Physical Transport Using the Technique of Explicit Differences 181 Geochemical Reactions 182 10.4 Calibration of the Model with Measured Values of DOC 182 10.5 Validation of the Model with Measured Values for Dissolved Iron 185 10.6 Discussion 187 10.7 References 189 11 Redox-Transport Modelling for the Oderbruch Aquifer Ekkehard Holzbecher, Christoph Horner, Gudrun Massmann, Asaf Pekdeger and Christoph Merz 191 11.1 Introduction 11.2 11.2.1 11.2.2 Site and Measurement Description 193 Hydrogeology 194 Water Chemistry 195 11.3 11.3.1 11.3.2 Modelling Concept 197 Reaction Model 198 Reactive Transport Simulation 202 11.4 11.4.1 11.4.1.1 11.4.1.2 Model Implementation Model Application Generic Precipitation/Dissolution Precipitation/Dissolution, Carbonate and Acid-Based Chemistry X 192 203 204 204 205 Contents 11.4.2 Parameterisation 206 11.4.2.1 Generic Precipitation/Dissolution 206 11.4.2.2 Precipitation/Dissolution, Carbonate and Acid-Based Chemistry 208 11.5 Results from Measurements and Modelling 210 11.6 Conclusions 11.7 References 213 12 Oxoanion Transport in Aquifers Containing Iron Hydroxide – Modelling of Column Experiments with PHREEQC2 Max Kofod, Verena Haury, Nandimandalam Janardhana Raju and Margot Isenbeck-Schröter 215 12.1 Introduction 12.2 Experimental Set-Up and Analytical Methods 216 12.3 12.3.1 12.3.2 12.3.3 12.3.4 Modelling of the Oxoanion Breakthrough Data Used Parameters Describing the Available Surface pH Buffering Cell Number and Diffuse Layer Options 12.4 12.4.1 12.4.2 12.4.3 Results Model Runs Using the Site Density of Amorphous Iron Hydroxide and Goethite Fitting the Site Densities pH Modelling 12.5 Summary and Conclusions 12.6 References 228 212 216 218 218 219 220 220 224 224 225 225 227 XI 16 Simulation of Two-Column Experiments on Anaerobic Degradation 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16.1 Introduction Aquifers polluted with aromatic hydrocarbons usually become anoxic and the majority of e.g BTEX compounds is degraded anaerobically A number of anaerobic microorganisms using different electron acceptors has been isolated in the past and the respective degradation pathways were studied (Heider et al., 1999) For toluene, it was found in all investigated cases that the primary reaction in the pathway is an addition of fumarate to the methyl group This reaction turned out to be a general mechanism for anaerobic bacteria to activate aromatic hydrocarbons containing methyl groups Also for o-xylene degradation this type of reaction was demonstrated However, in contaminated aquifers organisms are facing a complex mixture of hundreds of different compounds which could interfere with the degradation of individual substrates It is almost impossible to assess such in situ interactions analytically but modelling can be applied to describe complex interactions in situ We therefore started soil column experiments to investigate anaerobic degradation of mixtures of BTEX compounds and obtain a profound basis to model inhibition processes in the subsurface Soil columns with an undefined anaerobic microbial community were supplied with a mixture of the most prominent aromatic contaminants benzene, toluene, ethylbenzene and o,m,p-xylene Simulations were performed with a numerical reactive transport model to reproduce the observed concentration behaviour of the aromatic compounds which were subject to biodegradation and to test hypotheses on the controlling degradation mechanisms Furthermore, it was estimated which of the processes observed at the lab scale might be relevant for reactive transport modelling at field sites Two experiments out of a series of laboratory studies were selected for this purpose The first was conducted in order to study the microbially mediated degradation of benzene, toluene, m,p,o-xylene, and naphthalene with sulfate as terminal electron acceptor (Winter, 1997) No degradation could be observed for benzene, m,p-xylene and naphthalene and therefore the simulations were carried out for toluene and oxylene only This experiment will be referred to as Experiment The second experiment described in this paper (Experiment 2) was performed with special emphasis on the interaction of the toluene and o-xylene degradation processes 264 16.2 Set-Up of the Soil Columns 16.2 Set-Up of the Soil Columns 16.2.1 Experiment The glass columns used in this study were 38.5 cm long and had a diameter of 4,5 cm The first 30 cm of the columns were filled with contaminated soil material from the saturated zone of the “Testfeld Süd” site The remaining 8.5 cm were filled with water only Five sampling ports P1–P5 were installed along the column with P1–P4 located throughout the soil-filled part of the column and P5 at the water-filled section The concentrations recorded at P5 were used to represent the outlet concentrations of the column In addition, the water was sampled before entering the column A mineral medium containing a mixture of aromatic hydrocarbons and sulfate was pumped bottom to top through the columns for a period of 79 days The flow rate as well as the concentrations of solutes were not constant throughout the experiment The inflow concentrations for toluene varied between 17 µM and 154 µM, those for o-xylene between and 58 µM and those for sulfate between 230 µM and 1179 µM o-Xylene was injected throughout the whole period, while toluene injection lasted only up to day 52 of the experiment The mean flow rate was 450 mL d–1 16.2.2 Experiment In this experiment the same type of column was used as in Experiment Experiment was run for 109 days The mean flow rate was reduced to 290 mL d–1 for the initial 22 days and was further reduced to 150 mL d–1 for the remaining 87 days Concentrations were only recorded at the column inlet and outlet As in Experiment the inflow concentrations for o-xylene and toluene varied with time, with maximum concentrations of 374 µM of toluene and 478 µM of o-xylene o-Xylene was added more or less constantly throughout the experiment, while four periods of elevated toluene concentrations were applied Period lasted from the beginning of the experiment until day 11 with a mean toluene concentration of approximately 120 µM, and periods and lasted from day 53 until day 74 and from day 82 until day 101, respectively, with mean toluene input concentrations of approximately 50 µM Period from day 101 until the end of the experiment (day 109) was characterized by distinctly increased input concentrations for toluene of about 300 µM Between these periods inflow concentration for toluene were less than 10 µM The intermittent supply of toluene should promote the observation of the effects of toluene on o-xylene degradation 265 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16.3 Set-Up of the Numerical Model 16.3.1 Flow and Transport Model The multi-species reactive transport model TBC (Schäfer et al., 1998) was used for the simulation of both experiments The model column was discretised into 19 elements, each of 0.02 m length The temporal discretisation was 0.01 d The high temporal resolution was necessary to fulfill the stability and accuracy criteria required by the numerical model TBC employs a finite-difference approach for both flow and transport modelling The boundary conditions for the flow model were a fixed potential boundary at the column inflow and a prescribed flux boundary condition at the outflow side For the transport modelling a time-variant fixed concentration boundary at the inflow of the column was assumed The concentrations values at the inflow boundaries were taken from the measurements at the column inlet There were no independent measurements for the determination of the effective porosity and the dispersivity of the medium However, as the residence time of the solutes in the column (approx 4–5 hours) was much smaller than the time intervals between the concentration measurements (at least days), the solute movement inside the column could not be resolved and hence effective porosity and dispersivity values were of minor importance for the comparison of model results and observations A value of 15 % was assumed for the effective porosity and 0.01 m for the dispersivity These are reasonable assumptions for small scale transport in porous media 16.3.2 Degradation Model Microbial growth is the core of the biochemical reactions in the TBC model This growth is linked to substrate and electron acceptor concentrations via Monod-terms The back-coupling between microbial growth and reactive species consumption is performed via turnover coefficients and stoichiometric relationships The basic equations are exemplified for a single microbial group X, one substrate S and one electron acceptor E: 266 S E  dX  X = mmax   Ks + S KE + E  dt  growth microbial growth (1)  dX    = −mdec X  dt  dec microbial decay (2) 16.3 Set-Up of the Numerical Model dS  dX  = −Y  dt  dt  growth substrate consumption (3) dE dS = FS dt dt consumption of the electron acceptor (4) with X = concentration of microorganisms [M L–3] S = concentration of substrate [M L–3] E = concentration of electron acceptor [M L–3] µmax = maximum growth rate [T–1] µdec = constant decay rate [T–1] KS = Monod-constant [M L–3] Y = turnover coefficient [M M–1] FS = stoichiometric factor [M M–1] Inhibition is simulated with the help of inhibition terms IF (Kindred and Celia, 1989) An inhibition term is of the following form: IF = IC IC+ C I (5) with IC = inhibition constant [M L–3] CI = concentration of the inhibiting substance [M L–3] The value of IF is low if the concentration of the inhibiting substance CI is much larger than IC IF is multiplied to the right hand side of Eq (1) Therefore low values of IF decrease microbial growth On the other hand, low values of CI result in a value of IF close to unity and hence microbial growth is not limited by inhibition Equation (5) represents a non-competitive inhibition with a gradual transition between concentration ranges with and without inhibition The stiff system of non-linear differential equations describing microbial growth and the related species consumption is solved with the DGEAR-method, a special routine for stiff systems of equations (Hindmarsh, 1974) Transport and reaction are coupled by an iterative multistep-procedure More details on the TBC model can be found in Schäfer et al (1998) 267 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16.4 Model Calibration For the calibration of the model it was assumed that two different types of microorganisms co-existed in the column: one group that performs exclusively toluene degradation, and a second group which exclusively degrades o-xylene Both groups use sulfate as the only electron acceptor, the hydrocarbons as sole carbon and energy sources (heterotrophic growth) and they completely mineralize the organic carbon to carbon dioxide and water Thus the growth equations of the two groups comprise the maximum growth rates, Monod-terms for sulfate, toluene and o-xylene and an inhibition term for the o-xylene degraders The two microbial populations interact in a sense that o-xylene degradation is partly inhibited by the toluene degraders In a first attempt an inhibition term (Eq (5)) was used for the growth equation of the o-xylene degraders with toluene as the inhibiting substance The maximum growth rates and the turnover coefficients for toluene and oxylene used in the model were obtained from batch experiments with organisms that were isolated from the same soil columns (Meckenstock, 1999; Morasch et al., 2001) The constant decay rate was adjusted so that the microbial growth and decay dynamics derived from the observed reactive species behaviour could be reproduced The value of the Monod-constant for toluene was adapted to result in the observed residual toluene concentrations A low value was taken for the Monod-constant for sulfate so that sulfate concentrations will not limit microbial growth The stoichiometric factor for sulfate consumption was calculated from the underlying redox equations for a total oxidation of the substrates to CO2 Like for toluene degradation the Monod-constant for o-xylene was adapted in a way that the observed residual o-xylene concentrations could be reproduced in the model However, in the case of o-xylene the interaction with the inhibition term had to be taken into account as both the Monod-term and the inhibition term affect the calculated growth of the o-xylene degraders Therefore, only the o-xylene concentrations observed after day 52, when toluene was omitted from the medium and could no longer inhibit o-xylene degradation, were used for the determination of the Monod-constant It was somewhat surprising that the Monod-constant for o-xylene was lower than that for toluene, although the toluene degraders outperformed the o-xylene degraders However, the o-xylene decrease at the end of the experiment could not have been reproduced with a larger value of the Monod-constant The inhibition constant was adjusted on the basis of the o-xylene concentrations observed up to day 52 The parameters used in the calibrated model are displayed in Table 16.1 A comparison between observed and simulated concentrations for toluene, o-xylene, and sulfate at the column outlet is shown in Figure 16.1 A further important model parameter was the initial density of the microorganisms It turned out during model calibration that it was necessary to provide relatively large initial bacterial densities for the immediate inlet region of the column Otherwise the rapid degradation of the hydrocarbons inside the column could not have been reproduced The initial bacterial densitiy was µmol cell carbon per cm3 column mate- 268 16.4 Model Calibration 140 toluene [3M] 120 100 80 60 40 20 0 20 40 60 80 operation time [days] 70 o-xylene [3M] 60 50 40 30 20 10 0 20 40 60 80 20 40 60 80 1.5 sulfate [3M] 1.2 0.9 0.6 0.3 Figure 16.1: Comparison between observed and simulated concentrations for toluene (top), o-xylene (middle) and sulfate (bottom) at the column outlet for the simulations with toluene as inhibiting substance The symbols represent observed values, the solid lines show simulation results with reaction, and the dashed lines show input concentrations rial for the first cm of the column and 10–3 µmol cell carbon per cm3 for the rest of the column Identical values were used for both microbial groups Figure 16.2 provides the observed and calculated concentrations for toluene, o-xylene, and sulfate along the column for one observation time Obviously, both toluene and o-xylene degradation are already completed at the first observation point The assumed high initial bacterial densities at the column inlet seem reasonable if taken into account that the column used in Experiment was operated already for a longer period prior to this experiment so that microbial populations were already established 269 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation Table 16.1: Parameters of the biochemical model used for model calibration Parameter Meaning Value Toluene degraders o-xylene degraders d-1 0.3 d-1 Maximum growth rate µdec Decay rate 0.1 d-1 0.03 d-1 Ktol Monod-constant for toluene 143 µM toluene – Kxyl Monod-constant for o-xylene – 63 µM o-xylene Ksul Monod-constant for sulfate µM sulfate µM sulfate Ytol Turnover coefficient for toluene mol toluene/ mol cell carbon – Yxyl Turnover coefficient for o-xylene – 0.9 mol o-xylene/ mol cell carbon Fsul Stoichiometric coefficient for sulfate 3.7 mol sulfate/ mol toluene 3.5 mol sulfate/ mol o-xylene ICtol Inhibition constant for toluene – 23 µM toluene 150 60 120 50 o -xylene [3M] toluene [3M] µmax 90 60 30 40 30 20 10 0 0.1 0.2 0.3 0.4 column length [m] 0.1 0.2 0.3 0.4 column length [m] 1.5 1.2 sulfate [3M] 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 0.9 0.6 0.3 0 0.1 0.2 0.3 0.4 column length [m] Figure 16.2: Comparison between observed (symbols) and simulated concentrations (solid lines) for toluene (top, left), o-xylene (top, right), and sulfate (bottom) along the column for the observations from day 24 of Experiment 270 16.5 Testing of Hypotheses on the Interaction between Toluene and Xylene Degradation 16.5 Testing of Hypotheses on the Interaction between Toluene and Xylene Degradation 16.5.1 Inhibition of Xylene Degradation by a Substance other than Toluene During model calibration it was assumed that o-xylene degradation was limited directly by toluene Now it was investigated whether the inhibition could have been triggered also by a special inhibitor produced by the toluene degraders Such an inhibitor could be any excreted metabolite of toluene degradation which could interfere with the very similar o-xylene degradation pathway (e.g benzylsuccinate) The production of this substance in the model is directly proportionate to the gross growth of the toluenedegrading microorganisms The inhibition itself is again realized with the inhibition term shown before (Eq (5)) The inhibitor is a hypothetical substance which was not measured during the experiments Therefore, it cannot be compared to observations and its absolute concentration values are irrelevant The inhibition effect results from the ratio of the concentration of the inhibitor and the inhibition constant All other parameters of the biochemical model were adopted from the calibration run (cf Tab 16.1) Figure 16.3 shows that the observed o-xylene concentrations could have been reproduced with both assumptions concerning the inhibition mechanisms Hence, it cannot be decided if o-xylene degradation is inhibited directly by the presence of toluene or by any special inhibitor produced by the toluene degraders The inhibition mechanisms with toluene as inhibiting substance was used for further simulations, as the approach with a special inhibitor requires an additional species in the reaction model and therefore is numerically more demanding 70 o-xylene [3M] 60 50 40 30 20 10 0 20 40 60 80 operation time [days] Figure 16.3: Comparison between observed and simulated concentrations for o-xylene at the column outlet for the simulations with a special agent as inhibiting substance The symbols represent observed values, the solid line shows the actual results, and the dashed line shows the simulation results from the model calibration 271 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation 16.5.2 Competition for Sulfate In this numerical experiment it was tested whether the preferential degradation of toluene could be the result of a competition of the two microbial groups for their common electron acceptor sulfate To this end the inhibition term was removed from the model equations and the Monod constant for sulfate of the o-xylene degraders was strongly increased from µM to 250 µM This means that sulfate becomes a limiting substrate for the o-xylene degraders All other parameters remained unchanged (cf Tab 16.1) Figure 16.4 illustrates that the reproduction of the observed o-xylene concentrations was less successful than in the case with an inhibition mechanism The high value of 250 µM for the Monod-constant for sulfate was necessary to suppress o-xylene degraders during the period when both o-xylene and toluene were present in the column (up to day 52) For later times, however, the increased Monod-constant results in a too slow growth of the o-xylene degraders and hence in too large simulated o-xylene concentrations Obviously the higher sulfate concentrations available for the o-xylene degraders after day 52, when toluene was no longer supplied and hence toluene degraders were no longer active, did not sufficiently compensate the adverse effect of the increased Monod-constant on the o-xylene degraders The simulations suggest that sulfate is not a suitable parameter to control the interaction between toluene and o-xylene degraders 70 60 o -xylene [3M] 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 50 40 30 20 10 0 20 40 60 80 operation time [days] Figure 16.4: Comparison between observed and simulated concentrations for o-xylene at the column outlet for the simulations with competition for sulfate The symbols represent observed values, the solid line shows the actual results, and the dashed line shows the simulation results from the model calibration 272 16.5 Testing of Hypotheses on the Interaction between Toluene and Xylene Degradation 16.5.3 Degradation of Toluene and Xylene by a Single Microbial Group In the last numerical experiment concerning column Experiment the hypothesis was tested that toluene and o-xylene degradation could be performed by a single type of organism instead of two groups as it was assumed so far The population should use toluene preferentially if both toluene and o-xylene are present This mechanisms was realized in the model in a way that the bacteria are able to perform two different metabolic pathways The o-xylene degrading pathway is inhibited by the presence of toluene using the inhibition term from Eq (5) The inhibition constant was 0.5 µM of toluene All other parameters of the biochemical model were adopted from the preceding simulations (cf Tab 16.1) The simulation results for o-xylene are displayed in Fig 16.5 Up to day 52 both hypotheses on the degradation mechanisms (two different groups or one single group) yield comparable results However, from day 52 on, when toluene supply was stopped, the results are more convincing when using two different microbial groups in the model The reason for the differing results is that the o-xylene degrading population needs a certain time to develop when the inhibition by toluene is stopped This leads to a gradual increase in o-xylene degrading activity and a corresponding gradual decrease in o-xylene concentrations For the case of one single microbial group the microbial population already developed during toluene degradation These microorganisms abruptly switch from toluene to o-xylene degradation after day 52 and were therefore able to nearly completely degrade o-xylene from day 52 on The observations suggest that the first hypothesis assuming two different microbial populations is more likely However, at present there are no data available on how fast anaerobic bacteria and especially sulfate reducers can induce the synthesis of enzymes for BTEX degradation From cultivation studies it is known that some bacteria can need several months before degradation activities can be detected in 70 o -xylene [3M] 60 50 40 30 20 10 0 20 40 60 80 operation time [days] Figure 16.5: Comparison between observed and simulated concentrations for o-xylene at the column outlet for the simulations with one single microbial group The symbols represent observed values, the solid line shows the actual results, and the dashed line shows the simulation results from the model calibration (with two microbial groups) 273 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation microcosm studies Others (e.g denitrifying strains) can grow more rapidly Nevertheless, the finding from the modelling studies that two different types of organisms are responsible for the degradation is supported by the fact that two different sulfatereducing organisms have been isolated from the column which can grow only either on toluene (strain TRM1) or on o-xylene (strain OX39) (Meckenstock, 1999; Morasch et al., 2001) 16.6 Simulation of Experiment Discretisation and boundary conditions were taken from the simulations of Experiment with the flow rates and influx concentrations being adapted to the new situation The input concentrations for toluene and o-xylene for the 109 days of column operation are displayed in Figure 16.6 The only observations in Experiment besides the input values were the o-xylene concentrations at the column outlet For the initial model run the biochemical parameters described in Section 16.4 were used, that is two different microbial populations were assumed and o-xylene degradation was inhibited by the presence of elevated toluene concentrations (Figure 16.7) Obviously the modelled inhibition of o-xylene degradation was too weak from day 40 on In the simulations, o-xylene was nearly completely degraded, while the observations show increased o-xylene degradation around day 60 and day 80 To better adjust simulated o-xylene concentrations to the measurements the inhibition constant for toluene was reduced from 23 µM toluene as used in Experiment to µM Figure 16.8 reveals that the simulation results now correspond much better to the observations The lower inhibition constant results in a stronger inhibition of the 500 toluene, o -xylene [3M] 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 400 300 200 100 0 20 40 60 80 100 operation time [days] Figure 16.6: ment 274 Input concentration for toluene (solid line) and o-xylene (dashed line) in Experi- 16.6 Simulation of Experiment o-xylene degraders and therefore to increased simulated o-xylene concentrations at the column outlet In addition to reducing the inhibition constant, the initial density for the o-xylene degrading bacteria at the column inlet had to be reduced from the value of µmol cell carbon per cm3 column material as used in Experiment to 0.4 µmol cell carbon per cm3 Otherwise the higher o-xylene concentrations for the initial phase of Experiment could not have been reproduced Different initial bacterial concentrations are not unrealistic, as different columns (although of the same type) were used for the two experiments 500 o -xylene [3M] 400 300 200 100 0 20 40 60 80 100 operation time [days] Figure 16.7: Comparison between observed and simulated concentrations for o-xylene at the column outlet in Experiment using the biochemical parameters from Experiment (Table 16.1) The symbols represent observed values, the solid line shows the actual results, and the dashed line shows the input concentrations 500 o -xylene [3M] 400 300 200 100 0 20 40 60 80 100 operation time [days] Figure 16.8: Comparison between observed and simulated concentrations for o-xylene at the column outlet in Experiment with reduced inhibition constant The symbols represent observed values, the solid line shows the actual results, and the dashed line shows the input concentrations 275 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 At the end of Experiment the model results show increased o-xylene concentrations which are the consequence of the inhibition effect of the high toluene concentrations on o-xylene degraders at that time (cf Fig 16.6) The observations, however, not show increased o-xylene concentrations for the latest stages of Experiment A better correspondence between simulations and observations would have required an increase of the inhibition constants for later times of the experiment This hints on some kind of acclimatization effect of the o-xylene degraders towards elevated toluene concentrations However, it has to be considered that during this experiment microbial growth was limited by the substrates toluene and o-xylene only In cases where additional substances become limiting, e.g if the common electron acceptor sulfate becomes depleted, the toluene degraders might permanently outcompete the o-xylene degraders 16.7 Conclusions and Outlook Numerical simulations with a reactive transport model were performed for two column experiments on the anaerobic degradation of o-xylene and toluene The goal of the simulations was to test hypotheses on the interaction of toluene and o-xylene degradation processes The simulations and observations did not allow to distinguish whether xylene degradation was directly inhibited by the presence of toluene or whether the inhibition was controlled by a special inhibitor produced during toluene degradation Toluene itself could be inhibitory as it is structurally and chemically very similar to xylene and could therefore directly inhibit the degrading enzymes But also metabolites from toluene degradation could inhibit xylene degradation in later stages of the pathway as some metabolites are excreted in rather high concentrations into the medium The toxicity of such compounds has not been elucidated so far and thus toxic effects of metabolites other than pure inhibition of degradative enzymes cannot be excluded The results from another model run suggested that competition for the common electron acceptor of the two hydrocarbons, sulfate, was not controlling the degradation process during the experiments This was expected, however, because sulfate was added in excess and the medium homogeneously mixed which excludes local limitations due to heterogeneous distribution of substrates and electron acceptor Nevertheless, limiting supply of electron acceptor may influence the degradation of BTEX in the field to a large extent because inhibition effects as reported here might become dominant under competitive conditions Further simulations considered the question whether the two aromatic hydrocarbons are degraded by two different bacterial populations or the degradation is performed by a single organism which is able to use both substrates Here the comparison of model results with the observations rendered evidence that the first possibility (i.e two groups) is more likely 276 16.7 Conclusions and Outlook Finally, the model parameters determined during the simulations of the first experiment were used to simulate a second experiment which was conducted under comparable conditions, but with a special emphasis on the inhibition process With slightly modified parameters, the inhibition model proofed to be suitable for the simulation of the second experiment, too Observations from the final stage of the second experiment suggest that there is a certain acclimatization of the microbial population in the soil column to elevated toluene concentrations which leads to a nearly simultaneous degradation of both hydrocarbons The modelling of the soil column experiments revealed two features that are also of importance for the simulation of the degradation process on the field scale: • Degradation of toluene and o-xylene is probably performed by two different organisms Therefore, two microbial groups should be considered in field scale modelling of these degradation processes • There can be an inhibition of toluene or of one of its metabolites on o-xylene degradation which may especially become relevant if the two populations compete for certain limiting nutrients or electron acceptors Acknowledgement Many thanks to R Warthmann and C Winter, University of Konstanz, for performing the soil column experiments and B Schink for constant support Financial support was provided by the Deutsche Forschungsgemeinschaft through grants Da 222/2 and Schi 180/7 277 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 Simulation of Two-Column Experiments on Anaerobic Degradation 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16.8 References Heider, J.; Spormann, A M.; Beller, H R.; Widdel, F (1999): Anaerobic bacterial metabolism of hydrocarbons FEMS Microbiol Rev 22, 459–473 Herfort, M.; Ptak, Th.; Hümmer, O.; Teutsch, G.; Dahmke, A (1998): Testfeld Süd: Einrichtung der Testfeldinfrastruktur und Erkundung hydraulisch-hydrogeochemischer Parameter des Grundwasserleiters Grundwasser 3, 159–166 Hindmarsh, A C (1974): GEAR Ordinary differential equation system solver UCID–30001, Revision 3, Lawrence Livermore Laboratories, Livermore, Ca Kindred, J S.; Celia, M A (1989): Contaminant transport and biodegradation, Conceptual model and test simulations Water Resour Res 2, 1149–1160 Meckenstock, R U (1999): Fermentative toluene degradation in anaerobic defined syntrophic cocultures FEMS Microbiol Lett 177, 67–73 Morasch, B.; Annweiler, E.; Warthmann, R J.; Meckenstock, R U (2001): The use of a solid adsorber resin for enrichment of bacteria with toxic substrates and to identify metabolites: degradation of naphthalene, o-, and m-xylene by sulfate-reducing bacteria J Microbiol Meth 44, 183–191 Schäfer, D.; Schäfer, W.; Kinzelbach, W (1998): Simulation of reactive processes related to biodegradation in aquifers Structure of the three-dimensional reactive transport model J Contaminant Hydrol 31, 167–186 Winter, C (1997): Anaerober mikrobieller Abbau von BTX-Aromaten und Naphthalin in anoxischen Bodensäulen Staatsexamensarbeit, Universität Konstanz, pp 76 278 ...Deutsche Forschungsgemeinschaft Geochemical Processes Conceptual Models for Reactive Transport in Soil and Groundwater Edited by Horst D Schulz and Georg Teutsch Research Report Deutsche Forschungsgemeinschaft... Freising Weihenstephan; e-Mail: totsche@pollux.edv agrar.tu-muenchen.de Geochemical Processes: Conceptual Models for Reactive Transport in Soil and Groundwater Research Report Deutsche Forschungsgemeinschaft... 04318 Leipzig, e-Mail: richterm@rz.uni-leipzig.de 20 Geochemical Processes: Conceptual Models for Reactive Transport in Soil and Groundwater Research Report Deutsche Forschungsgemeinschaft (DFG)