Reactive transport modeling of subsurface flow CWs using the. HYDRUS wetland module. Content. Introduction The HYDRUS software a d e Sc e ces, e a

Transcription

1 Reactive transport modeling of subsurface flow CWs using the a d e Sc e ces, e a HYDRUS wetland module Günter Langergraber 1 and Jirka Šimůnek 2 1 Institute of Sanitary Engineering and Water Pollution Control 2 Environmental Sciences, University of California Riverside 4th International Conference HYDRUS Software Applications to Subsurface Flow and Contaminant Transport Problems March 213, Prague, Czech Republic Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 1 Content Treatment wetlands Models for wetlands The Wetland Module of HYDRUS Simulation results 1. Vertical flow constructed wetlands 2. Batch-fed CWs (Master thesis Tamás Pálfy, BOKU) 3. HF beds with peak loads (Poster Rizzo and Langergraber) 4. HF beds for tertiary treatment (Poster Sanchez-Ramos and Langergraber) 5. Nitrate dynamics in a rural headwater catchment: measurements and modelling (Poster Smethurst and Langergraber) Summary and conclusion Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 2 The HYDRUS software a d e Sc e ces, e a The HYDRUS wetland module a d e Sc e ces, e a Software Package for simulating the two- and three-dimensional movement of water, heat, and multiple l solutes in variably-saturated t media HYDRUS numerically solves the Richards equation for saturated/unsaturated water flow and the convection-dispersion equation for heat and solute transport. Graphical User Interface Version 2 released in May 211 HYDRUS Wetland module (Langergraber and Šimůnek, 26, 211) Šimůnek et al., 211, HYDRUS Technical and User manuals, version 2, Langergraber and Šimůnek (211): HYDRUS Wetland Module, Version 2. Hydrus Software Series 4, 56p. Langergraber and Šimůnek (26): The multi-component reactive transport module CW2D for CWs for HYDRUS. Hydrus Software Series 2, 72p. Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 3 CW2D, 21 Langergraber PhD Version 1 (26): CW2D Version 2 (211): CWM1+CW2D Šimůnek et al., 28, Vadoze Zone J 7(2), Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 4

2 Types of wetlands (Fonder and Headley, 21) a d e Sc e ces, e a Natural wetlands are those wetland areas that exist in the landscape due to natural processes rather than having been created either directly or indirectly as a result of anthropogenic influences. Constructed wetlands are man-made systems that are designed to mimic many of the conditions and/or processes that occur in natural wetlands. Purpose of constructed wetlands: Restored wetlands: Areas which were formerly natural wetlands Created wetlands: Non-wetland areas which have been converted Treatment wetlands: Artificially created wetland systems designed to provide a specific water treatment function Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 5 Types of wetlands a d e Sc e ces, e a Fonder and Headley (21) Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 6 Treatment wetlands (Fonder and Headley, 21) a d e Sc e ces, e a 3 characteristics can be identified which are typical of all TWs: The presence of macrophytic vegetation ti that t typically grows within natural wetlands; The existence of water-logged or saturated substrate conditions for at least part of the time; and The inflow of contaminated waters with constituents that have to be removed. Treatment wetlands (Fonder and Headley, 21) a d e Sc e ces, e a 3 characteristics can be identified which are typical of all TWs: The presence of macrophytic vegetation ti that t typically grows within natural wetlands; The existence of water-logged or saturated substrate conditions for at least part of the time; and The inflow of contaminated waters with constituents that have to be removed. Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 7 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 8

3 TW main types a d e Sc e ces, e a HF CW Kyjov, Czech Republic Picture: Jan Vymazal surface flow (free water surface) subsurface flow horizontal flow vertical flow with intermittent loading Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 9 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber Models for wetlands a d e Sc e ces, e a In wetlands a large number of physical, chemical, and biological processes are active in parallel and mutually influence each other. Therefore wetlands are complex systems and for a long time have been often considered as "black boxes". Also models for wetlands for a long time have been using "black box" approaches only, i.e. the processes in wetlands have not been considered in detail. Still today most "models" for wetlands are using a "black box" approach The number of models describing processes in wetlands in detail is limited. Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 11 Models for wetlands a d e Sc e ces, e a "Simple" models ("black-box" approach) correlations between influent and effluent concentrations ti first-order rate equations (used for design of TWs) artificial neural network models etc. Influent Effluent Data from experiments needed to derive model equations Trang et al. (29) Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 12

4 Models for wetlands a d e Sc e ces, e a Process-based models Mathematical ti model equations based processes in wetlands (with various degree of complexity) - includes balance equations for energy, mass, charge, dm dc V Typical mass QIN CIN Q COUT r V balance equation dt dt OUT Data are used for calibration and validation of model Better prediction is possible using this models should be better applicable for design Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 13 x Models describing wetland processes Processes in wetlands a d e Sc e ces, e a For a WETLAND MODEL a number of different processes has to be considered: The flow model (describing water flow) The transport model (describing transport of constituents as well as adsorption and desorption processes) The biokinetic model (describing biochemical i transformation ti and degradation processes) The influence of plants (growth, decay, decomposition, nutrient uptake, root oxygen release, etc.) The description of clogging processes Physical re-aeration Langergraber, G., Rousseau, D.P.L., PL García, J., Mena, J. (29): CWM1 - A general model to describe biokinetic processes in subsurface flow constructed wetlands. Water Sci Technol 59(9), Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 14 HYDRUS wetland module Processes available a d e Sc e ces, e a For a WETLAND MODEL a number of different processes has to be considered: The flow model (describing water flow) The transport model (describing transport of constituents as well as adsorption and desorption processes) The biokinetic model (describing biochemical i transformation ti and degradation processes) HYDRUS wetland module The influence of plants (nutrient uptake, root oxygen release) The influence of plants (growth, decay, decomposition) The description of clogging processes not available in HYDRUS Physical re-aeration HYDRUS wetland module Wetland module General description a d e Sc e ces, e a HYDRUS wetland module developed d for SSF TWs includes two biokinetic model formulations: 1. CW2D (Langergraber g and Šimůnek, 25) 2. CWM1 (Constructed Wetland Model #1) (Langergraber et al., 29). Model CW2D CWM1 Processes Aerobic and anoxic (9) Aerobic, anoxic and anaerobic (17) Constituents Oxygen, organic matter, nitrogen and phosphorus (12) Oxygen, organic matter, nitrogen and sulphur (16) Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 15 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 16

5 15 1 Simulation Results 1 Pilot-scale subsurface VF CW for wastewater treatment a d e Sc e ces, e a INLET 6 main layer intermediate layer drainage layer OUTLET Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 17 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 18 Simulation Results 1 Pilot-scale subsurface VF CW for wastewater treatment a d e Sc e ces, e a Flow simulations Simulation Results 1 Pilot-scale subsurface VF CW for wastewater treatment a d e Sc e ces, e a Microbial biomass Effluent ra ate (ml/min) 6 Measured data 5 Measured ks + porosity Langergraber (23) Time (hours) 3% -1cm simulated (2cm) 25% simulated (5cm) simulated (1cm) r content water 2% 15% 1% measured (-1cm) 5% % time after feeding [h] FUM, MDC / gdw soil) SIR, ATP, C-F (µg BM-C / SIR: microbial BM-C ATP: microbial BM-C C-Fumigation: microbial BM-C MDC: bacterial BM-C 24 1 N-Fumigation: microbial BM-N Depth (cm) M gdw soil) N-FUM (µg BM-N / g s COD (mg/g DW) Microbial biomass M 7' 6' 5' 4' 3' 2' 1' Calculated from biomass C (SIR) Simulated Filter depth (cm) Tietz et al., 27, Sci Total Environ 38(1-3), Langergraber et al., 27, Water Sci Technol 56(3), Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 19 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 2

6 Simulation Results 1 Outdoor CW (Ernsthofen, NÖ) for treatment of wastewater a d e Sc e ces, e a Langergraber (27): Sci Total Environ 38(1-3), parallel beds surface area 2 m² each organic loading of 2, 27, 4 g COD/(m².d) resp. 4, 3, 2 m²/pe Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 21 Simulation Results 1 Outdoor CW (Ernsthofen, NÖ) for treatment of wastewater a d e Sc e ces, e a low (L/h) Effluent fl 1 Measured effluent flow Simulated effluent flow Measured cumulative effluent Simulated cumulative effluent Langergraber (27): Sci Total Environ 38(1-3), Time (h) effluent (L) Cumulative Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 22 Simulation Results 1 Outdoor CW (Ernsthofen, NÖ) for treatment of wastewater a d e Sc e ces, e a 7 Simulation Results 2 Master thesis Tamás Pálfy a d e Sc e ces, e a NH4-N (mg/l) CSB, NH4-N simulated COD simulated NH4-N measured COD measured Goal: to verify the implementation of CWM1 in the HYDRUS Wetland Module Uses column experiment data made by a research group at Montana State University as reference outcome Compare with simulation results published by Mburu et al. (212, Ecol Eng 42, ) CWM1 implemented in Aquasim Temperature ( C) After adaptation of temperature dependencies for K ANs,NH4 and K X Langergraber (27): Sci Total Environ 38(1-3), Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 23 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 24

7 May 31, 27 Simulation Results 2 Master thesis Tamás Pálfy a d e Sc e ces, e a Experiments at MSU D=2cm, h=5 cm batch-operated columns filled with pea gravel Synthetic domestic wastewater 49 ± 4.3 mg/l COD 14 ±.5 mg/l SO 4 -S 4 ± 1.2 mg/l TN 8 ±.3 mg/l PO 4 Operated at 12, 16, 2 and 24 C 2 days incubation Sedge, bulrush and cattail plus unplanted column for each run Simulation Results 2 Master thesis Tamás Pálfy a d e Sc e ces, e a Simulation results, 24 C unplanted Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 25 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 26 Simulation Results 3 Poster: A. Rizzo* and Langergraber a d e Sc e ces, e a Simulation of peak loads for horizontal flow CWs - A first step towards modelling stochastic behaviour of HF CW Objective Compare simulation results for a HF CW subjected to sudden peak loads s [Galvão and Matos, 212, Ecol Eng 49, ] Data from laboratory experiment using synthetic wastewater COD load was applied Lines A 11.4 g/m²/day for 3.5 months, then +22% load for 2 week Lines B 5.3 g/m²/day for 3.5 months, then +7% load for 2 week * Environment, Land and Infrastructure Engineering (DIATI), Politecnico di Torino, Italy Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 27 Simulation Results 3 Poster A. Rizzo* and Langergraber a d e Sc e ces, e a Main assumptions parameters CWM1 biokinetic model [Langergraber, G., Šimůnek, J., 212] rate constant of lysis of fermenting bacteria (X FB ).5 Fraction of inert particulate (X I ) generated in biomass lysis.1 no O2 inflow, no plants maximum time step.1 d (ca. 1.5 min) new time step check to limit numerical disturbance (code correction was implemented in the new HYDRUS version) COD in,tot fractionation for syntheic wastewater S F =62% COD in,tot ; S A =1% COD in,tot ; S I =3% COD in,tot ; X S =2% COD in,tot ; X I =5% COD in,tot Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 28

8 Simulation Results 3 Poster A. Rizzo* and Langergraber a d e Sc e ces, e a Results 1 COD outflows AE COD,out Rem COD,meas Rem COD,sim adequately simulated [%] [%] [%] A B Simulation Results 3 Poster A. Rizzo* and Langergraber a d e Sc e ces, e a Results 1 NH4+ outflows not adequately simulated AE COD,out [%] A 83 B 6 A B A B Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 29 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 3 Simulation Results 4 Poster: D. Sanchez-Ramos* and Langergraber a d e Sc e ces, e a Modelling horizontal flow constructed wetlands treating effluents of wastewater treatment plants Objective Tablas de Daimiel National Park (Spain): Floodplain wetland (1928 ha) connected with an important t aquifer (55 km²) Intensive pumping disconnection drying Treated Sewage Effluents from several WWTPs in the surroundings as possible solution for maintenance of TDNP until aquifer recovery Direct use of TSE unsafe CW systems to polish WWTP effluents * School of Civil Engineering, University of Castilla-La Mancha, Ciudad Real, Spain Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 31 Simulation Results 4 Poster: D. Sanchez-Ramos* and Langergraber a d e Sc e ces, e a HF CW design: Treatment cells: 23 m (length) x 22 m (width) x.6 m (depth) HLR = 19 cm d -1 Implementation in HYDRUS 2D finite it element mesh (25 rows x 25 columns) Atmospheric boundary condition at the top of the mixing zone; constant head boundary condition at the bottom of the right end of the bed (effluent boundary) Influence of plants (Phragmites australis): Transpiration rate: 7.4 mm d -1 O 2 release: 5 g m -2 d -1 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 32

9 Simulation Results 4 Poster: D. Sanchez-Ramos* and Langergraber a d e Sc e ces, e a Influent concentrations: 125 mg O2 L -1 COD (mainly as slowly biodegradable particulate COD, XS) and 15 mg N L -1 ammonium (SNH) CWM1 parameters changed Parameter CWM1 original value Simulation value Variation (%) μ H (d -1 ) 6 2,5-58,33 b H (d -1 ) 4,4 4,4 K OH (mg O 2 /l),2,4 1 K h (d -1 ) ,33 f BM,XI (g COD XI /g COD BM ) 1,1 2,2-8 Simulation Results 4 Poster: D. Sanchez-Ramos* and Langergraber a d e Sc e ces, e a Simulation results Better treatment efficiency than expected in the design: - 66% COD removal - 39% NH 4 removal High HRT (19 cm d -1 ) important O 2 and COD supply in the inlet The high development of X H has been the main difficulty on the simulation Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 33 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 34 Simulation Results 5 Poster: P. Smethurst* and Langergraber a d e Sc e ces, e a Nitrate dynamics in a rural headwater catchment: measurements and modelling Willow Bend Farm Paired-Catchment Experiment 27 Study Catchment: Willow Bend Farm, Forsters Rivulet Treated Catchment 1 ha 38% native forest 62% pasture Control Catchment 4 ha 1% native forest 99% pasture Smethurst PJ, Petrone KC, Langergraber G, Baillie CC, Worledge D, Nash D (213) Nitrate dynamics in a rural headwater catchment: measurements and modelling. Hydrological Processes (in press). DOI: 1.12/hyp.979 * CSIRO Ecosystem Sciences, Hobart, Tasmania Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 35 Insert presentation title Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber

10 SMZ plantation at 4 years of age Simulation Results 5 Poster: P. Smethurst* and Langergraber a d e Sc e ces, e a Simulation Results 5 Poster: P. Smethurst* and Langergraber a d e Sc e ces, e a HYDRUS Wetland Modelling: simulation of riparian buffer N dynamics Important Capabilities: Denitrification Other organic C, N and P pools Rooting depth N uptake Rainfall, runoff, groundwater, and seepage Spatial and temporal flexibility Z X Main novel steps: 1. External spread-sheet: quick-flow/slow-flow analysis 2. HYDRUS: route slow-flow through HYDRUS as infiltration 3. External spread-sheet: recombine seepage as base-flow with quick-flow to estimate t stream-flow Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 37 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 38 Simulation Results 5 Poster: P. Smethurst* and Langergraber a d e Sc e ces, e a Water and Nitrate: hillslope modelling success at hourly and annual scales Simulation Results 5 Poster: P. Smethurst* and Langergraber a d e Sc e ces, e a Water and Nitrate: hillslope modelling success at hourly and annual scales Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 39 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 4

11 Summary Different applications of the biokinetic models CW2D and CWM1 Biokinetic model CW2D (Langergraber and Šimůnek, 25) CWM1 (Langergraber et al., 29) Type of CW VF CWs VF and HF CWs Low loaded HF beds Processes Modeling retention of Modeling anaerobic phosphorus in CWs processes Modeling nitrification Modeling transport and as a 2-step process fate of sulphur Conclusions The following conclusions can be drawn: Version 2 of the HYDRUS wetland module is the only publicly available implementation of the Constructed Wetland Model N 1 (CWM1). It is essential for modelling SSF CWs that fixed bacteria can be simulated. Since CWM1 is able to describe anaerobic processes, it is more suitable for modelling HF CWs. The influence of wetland plants on various biochemical transformation and degradation d processes due to release of oxygen by plant roots in a HF bed is significant and therefore has to be considered. More experience still has to be gained in using the CWM1 biokinetic model. Langergraber and Šimůnek (212, Vadoze Zone J 11(2), doi:1.2136/vzj Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 41 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 42 Needs for improvement 1 (view of the Wetland module users) a d e Sc e ces, e a New time step adjustment linked with solute concentrations calculated in the wetland module Adjust default parameters for biokinetic models based on recent experiences Make limited outflow function again Particle transport and influence on hydraulic properties It is not possible to use physical non-equilibrium and chemical nonequilibrium at the same time (presentation Ania) Preferential flow is not possible while using the wetland module Implementation of flexible biokinetic model, i.e. addition of substances and processes to the wetland module should be possible by users Needs for improvement 2 (view of the Wetland module users) a d e Sc e ces, e a Overland flow needs to be simulated by pre- and post-hydrus processing (hillslope application). For hillslope applications of HYDRUS, with or without the wetland module, it would be very useful to include outputs of pools and fluxes of both water and nitrate on a per area basis. Only one root type can be specified s specified, wherever they are placed in the simulation domain (hillslope application). Nutrient uptake is simulated only as the mass flow component. For nutrients that are taken up mainly via diffusive processes, e.g. ammonium, phosphorus and potassium, simulated uptake is likely to be greatly underestimated using HYDRUS. A version of HYDRUS is now available that includes active and passive uptake (Šimůnek and Hopmans 29). Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 43 Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 44

12 Contact Dr Guenter Langergraber (BOKU) Water, Atmosphere and Environment Institute of Sanitary Engineering and Water Pollution Control Muthgasse 18, A-119 Vienna, Austria Tel.: +43 () , Fax: +43 () Institute of Sanitary Engineering and Water Pollution Control I Günter Langergraber 45