acknowledgments - virginia tech · 2020-01-18 · acknowledgments . the author wishes to express...
TRANSCRIPT
ACKNOWLEDGMENTS
The author wishes to express sincere appreciation and gratitude to Professor Mark A. Widdowson for
his assistance, guidance, and support in the preparation of this dissertation. As much as I look at Dr.
Widdowson as a great advisor, as much as I look at him a source of inspiration. Yet he did everything
one can possibly expect from an advisor, he generated as many new ideas as I could handle, spent all
the hours of the world discussing whatever technical issue was in my mind, made sure I had the best
environment to work in, gave me many opportunities to present my thoughts, and had a warm heart
and support in difficult moments.
In addition, special thanks to Mr. Eduardo Mendez III for his help preparing and testing the
SEAM3D-PUP code. Ed was a close friend whose efforts are greatly appreciated.
The committee members Dr. John T. Novak, Dr. Thomas J. Burbey, Dr. G.V. Loganathan, and Dr.
Conrad D. Heatwole are the ideal images of university professors. I learned a lot from them during my
study and I was impressed by their ability in teaching and research that I hoped to have them in my
committee and my wish came true. I thank them for the insights and suggestions throughout the
research.
I also thank Dr. Randal Dymond, VT professor and director of CGIT for his help and support and
giving me the opportunity to create and teach in my beloved department of EWR, VT. My deepest
gratitude to Prof. Bill Knocke for supporting me during my last two years of study. Thanks to all my
colleagues in the department of civil and environmental engineering.
Special thanks to my wife, Mona, for her love and patience, for putting up with late nights, and for
taking care of my sons most of the time. Thanks for my parents for their sincere prayers for me.
Thanks for every one contributed to this dissertation by teaching me even the smallest thing in my life,
for all of those people I hope that I can repay you by being honest and helpful to all human being.
iv
TABLE OF CONTENTS
CHAPTER 1......................................................................................................................................... 1 INTRODUCTION ................................................................................................................................. 1 1-1 BACKGROUND.............................................................................................................................. 1
Processes known to degrade/remove contaminates..................................................................... 2 Application of Phytoremediation to Groundwater Contaminants............................................... 3 Phreatophytes................................................................................................................................ 6 Use of Trees (Phreatophytes) in Phytoremediation ..................................................................... 7
1-2 OBJECTIVES.................................................................................................................................. 8 1-3 ORGANIZATION OF THE DISSERTATION........................................................................................ 8
CHAPTER 2....................................................................................................................................... 10 LITERATURE REVIEW..................................................................................................................... 10 2.1 INTRODUCTION ........................................................................................................................... 10 2.2 PHYTOREMEDIATION .................................................................................................................. 13
2.2.1 Applicability of Phytoremediation..................................................................................... 13 2.2.2 Hyperaccumulators............................................................................................................ 14 2.2.3 Poplar Trees....................................................................................................................... 14 2.2.4 Phytoremediation Mechanisms.......................................................................................... 16
2.3 MODELS FOR PHYTOREMEDIATION PROCESSES......................................................................... 20 2.3.1 Plant Uptake....................................................................................................................... 20
2.3.1.1 Local point-/field-scale models .................................................................................. 21 2.3.1.2 Diffuse sink root models............................................................................................. 21 2.3.1.3 Large-scale atmospheric modeling............................................................................. 25 2.3.1.4 Models for direct Transpiration.................................................................................. 26 2.3.1.5 Equilibrium Models for Transpiration ....................................................................... 34
2.3.2 Root Sorption...................................................................................................................... 37 2.3.2.1 Equilibrium Concentrations........................................................................................ 39 2.3.2.2 Equilibrium Plant uptake Models............................................................................... 41 2.3.2.3 Sorption/desorption Kinetics ...................................................................................... 43 2.3.2.4 Root Concentration Factor, RCF................................................................................ 44
2.3.3 Rhizosphere Biodegradation.............................................................................................. 48 2.4 RESEARCH ON PHYTOREMEDIATION .......................................................................................... 48
2.4.1 Modeling Phytoremediation: Previous Work.................................................................... 49 2.5 PHYTOREMEDIATION TECHNICAL CONSIDERATIONS................................................................. 60
2.5.1 Advantages of Phytoremediation....................................................................................... 62 2.5.2 Limitations of Phytoremediation ....................................................................................... 62 2.5.3 Costs of Phytoremediation................................................................................................. 62
2.6 RESEARCH DEFICIENCIES ........................................................................................................... 64 2.7. RESEARCH AIMS........................................................................................................................ 64
CHAPTER 3....................................................................................................................................... 66 MODEL DEVELOPMENT.................................................................................................................. 66
v
3.1 CONCEPTUAL MODEL................................................................................................................. 66 3.2 MATHEMATICAL MODEL............................................................................................................ 67
3.2.1 Direct Uptake ..................................................................................................................... 67 3.2.2 Root Sorption...................................................................................................................... 69
3.3 MODEL IMPLEMENTATION ......................................................................................................... 70
CHAPTER 4....................................................................................................................................... 73 MODEL TESTING AND VERIFICATION ........................................................................................... 73 4.1 VERIFICATION OF THE PLANT UPTAKE PACKAGE...................................................................... 73 4.2 PLANT UPTAKE........................................................................................................................... 73
4.2.1 Closed System Model – Single Stress Period .................................................................... 73 4.2.2 Closed System Model – Multiple Stress Periods............................................................... 74 4.2.3 Flow and Transport with Direct Uptake ........................................................................... 75
4.3 ROOT SORPTION ......................................................................................................................... 76 4.3.1 Flow and Transport with Root Sorption (f = 1.0)............................................................. 76 4.3.2 Flow and Transport with Root Sorption (f < 1.0)............................................................. 77 4.3.3 Flow and Transport with Root Sorption (f = 1.0) and Aquifer Sorption ......................... 78 4.3.4 Flow and Transport with Spatially-Variable Root Sorption (f = 1.0).............................. 78
4.4 DIRECT UPTAKE AND ROOT SORPTION ...................................................................................... 79 4.4.1 Flow and Transport with Plant Uptake and Root Sorption (in the ET area only)........... 79 4.4.2 Flow and Transport with Spatially Distributed RCF and Plant Uptake.......................... 80
4.5 CONCLUSIONS............................................................................................................................. 80
CHAPTER 5..................................................................................................................................... 112 SIMULATION OF A PHYTOREMEDIATION SYSTEM USING SEAM3D-PUP ............................... 112 5.1 INTRODUCTION ......................................................................................................................... 112 5.2 MODEL DESCRIPTION ............................................................................................................... 114 5.3 RESULTS AND DISCUSSIONS ..................................................................................................... 118
5.3.1 Initial Test Case ............................................................................................................... 118 5.3.2 Effect of ET area (WET and LET) on contaminant mass removal .................................... 121 5.3.3 Effect of ET Area on Plume Concentration..................................................................... 124 5.3.3.1 Radioactive decay or biodegradation .......................................................................... 136 5.3.4 Effect of ET area and TSCF on mass-flux....................................................................... 138
5.4 EFFECT OF GROUNDWATER FLUX AND ET FLUX RATES .......................................................... 144 5.4.1 Effect of Aquifer In-Flux/ Out-Flux on Mass Removal ................................................... 147 5.4.2 Effect of aquifer in-flux/ out-flux on plume concentration.............................................. 151 5.4.3 Effect of Aquifer In-Flux/Out-Flux on Average Solute Mass-Flux................................. 154
5.5 EFFECT OF DIVIDING THE ET AREA INTO TWO HALVES.......................................................... 158 5.6 EFFECT OF REMOVING THE SOURCE......................................................................................... 161 5.7 PHYTOREMEDIATION SYSTEM DESIGN METHODOLOGY.......................................................... 166
5.7.1 Design Example 1 ............................................................................................................ 173 5.7.2 Design Example 2 ............................................................................................................ 175
CHAPTER 6..................................................................................................................................... 176 ALTERNATIVE MODEL FOR SEAM3D-PUP............................................................................... 176 6.1 INTRODUCTION ......................................................................................................................... 176
6.1.1 Plant Uptake – Power Relationship ................................................................................ 176
vi
6.1.2 Plant Uptake – Plant Concentration Capacity ............................................................... 178 6.1.3 Objective........................................................................................................................... 179
6.2 MATHEMATICAL MODELS........................................................................................................ 180 6.2.1 Freundlich Isotherm (Power Function) .......................................................................... 180 6.2.2 Langmuir Sorption Isotherm (Plant Tolerance) ............................................................. 181
6.3 MODEL VERIFICATION ............................................................................................................. 183 6.3.1 Freundlich (ISO=2) Verification..................................................................................... 183 6.3.2 Langmuir (ISO=3) Verification....................................................................................... 187
6.4 ALTERNATIVE MODEL APPLICATIONS, PCE SIMULATION....................................................... 193
CHAPTER 7..................................................................................................................................... 197 CONCLUSIONS AND RECOMMENDATIONS ................................................................................... 197 RECOMMENDATIONS FOR FUTURE RESEARCH............................................................................... 199
CHAPTER 8..................................................................................................................................... 201 INPUT INSTRUCTIONS ................................................................................................................... 201
SEAM3D MODEL INPUT............................................................................................................. 201 GENERAL INFORMATION ................................................................................................................ 201
Types of Input ............................................................................................................................ 201 Array Readers............................................................................................................................ 202 Units........................................................................................................................................... 203
INPUT INSTRUCTIONS ..................................................................................................................... 203 Input Instructions for the Plant Uptake Transport Package.................................................... 203
RCF Notes ............................................................................................................................. 205 TSCF Notes ........................................................................................................................... 206
FREUNDLICH, ISO=2...................................................................................................................... 208 LANGMUIR, ISO=3......................................................................................................................... 208
BIBLIOGRAPHY .................................................................................................................................. 209 APPENDIX A....................................................................................................................................... 220
AUXILIARY FIGURES AND TABLES FROM CHAPTER 5.................................................................... 220 VITA.................................................................................................................................................. 251
vii
LIST OF FIGURES
Figure 1.1. Schematic of phytoremediation processes. ...................................................................................... 1 Figure 1.2. Superfund Remedial Actions: Source Control Treatment Projects (FY 1982 - 2002). .......... 5 Figure 2.1. Signs used in Love Canal community of New York....................................................................11 Figure 2.2. Solute fate in plants.............................................................................................................................20 Figure 2.3. A schematic overview of the SWAP model system. ....................................................................25 Figure 2.4. Volumetric evapotranspiration, QET, as a function of head, h, in a cell where d is the
extinction depth, and hs is the ET surface elevation...........................................................................27 Figure 2.5. Representation of evapotranspiration in MODFLOW...............................................................28 Figure 2.6. Plan view of model grid (left) and cross section of model grid (right) used in evaluating
aquifer properties effect on phytoremediation effectiveness............................................................28 Figure 2.7. Effect of growing season duration on minimum plantation area for capture.........................30 Figure 2.8. Effect of aquifer anisotropy on minimum plantation area for capture. ...................................30 Figure 2.9. Effect of plume width on minimum plantation area for capture. .............................................31 Figure 2.10. Effect of water table, and root depth on ET rate.......................................................................31 Figure 2.11. Relationship between the translocation of chemicals to barley shoots following uptake by
roots over 24 h (expressed as the Transpiration Stream Concentration Factor, TSCF) and their 1-octanol/water partition coefficient (as log Kow); ο, O-methylcarbamoyloximes; ×, substituted phenylureas............................................................................................................................35
Figure 2.12. Equilibrium modeling levels. ..........................................................................................................42 Figure 2.13. Relationship between the uptake of chemicals by plant roots (expressed as the Root
Concentration Factor, RCF) from nutrient solution at 24 h and their 1-octanol/water partition coefficient (as log Kow) for O-methylcarbamoyloximes and substituted phenylureas. ................48
Figure 2.14. Decision tree for phytoremediation. .............................................................................................61 Figure 3.1. Conceptual model for the two main mechanisms simulated using the SEAM3D Plant
Uptake Package. ........................................................................................................................................71 Figure 3.2. SEAM3D-PUP flowchart. ................................................................................................................72 Figure 4.1. Schematic of a closed system model for testing the direct uptake feature using the
SEAM3D-RDP. ........................................................................................................................................96 Figure 4.2. Simulated dissolved concentration and mass removed by direct uptake versus time from
SEAM3D-PUP and SEAM3D-SSM with TSCF = 1.0 for the closed-system, single stress period model in Figure 3.1. .....................................................................................................................97
Figure 4.3. Simulated dissolved concentration (top) and mass removed by direct uptake (bottom) versus time using SEAM3D-PUP for the closed-system model in Figure 3.1 for the range of TSCF values, varying from 0 to 1.0.......................................................................................................98
Figure 4.4. Simulated dissolved concentration and mass removed by direct uptake versus time from SEAM3D-PUP and SEAM3D-SSM with TSCF = 1.0 for the closed-system model in Figure 4.1 with two stress periods with variable rates of evapotranspiration (top). .................................99
Figure 4.5. Simulated dissolved concentration (top) and mass removed by direct uptake (bottom) versus time using SEAM3D-PUP for a four stress period, closed-system model in Figure 3.1 for the range of TSCF values, varying from 0 to 1.0. ......................................................................100
viii
Figure 4.6. Conceptual model for case study 3.1.3, flow and transport with direct uptake in the ET area (no root sorption; TSCF is T, and RCF is F). Three observation points are noted: (i, j, k) = (24, 45, 1), (24, 50, 1), and (24, 56, 1). .................................................................................................101
Figure 4.7. Mass removal by direct uptake versus time using SEAM3D-PUP and SEAM3D-SSM for a one-stress period, flow-system model shown in Figure 4.6 for the range of TSCF values, varying from 0.0 to 1.0...........................................................................................................................102
Figure 4.8. Concentration versus time using SEAM3D-PUP and SEAM3D-SSM for a one-stress period, flow-system model shown in Figure 4.6 (test case 4.1.3) for the three observation points (top), and for the middle observation point for the range of TSCF values, varying from 0.0 to 1.0 (bottom). ............................................................................................................................................103
Figure 4.9. Concentration versus time for the middle observation point, (i, j, k) = (24, 50, 1) using SEAM3D-PUP (top), and comparing it with SEAM3D-RCT (bottom) for case study 4.2.1. 104
Figure 4.10. Mass removal versus time for the middle observation point, (i, j, k) = (24, 50, 1) using SEAM3D-PUP and comparing it with SEAM3D-RCT for case study 4.2.1. ............................105
Figure 4.11. Hydraulic head distribution for r =24 (top), and concentration versus time for the middle observation point, (i, j, k) = (24, 50, 1) using SEAM3D-PUP, and comparing it with SEAM3D-RCT (bottom) for case study 4.2.2. .....................................................................................................106
Figure 4.12. Concentration versus time for the middle observation point, (i, j, k) = (24, 50, 1) using SEAM3D-PUP where 50% of the retardation is due to plant roots and 50% is due to soil matrix, and comparing it with SEAM3D-RCT where 100% of the retardation is due to soil matrix for case study 4.2.3.....................................................................................................................107
Figure 4.13. Screen capture for the results of R in case study (4.2.3.1) showing R=2.0 in the roots cells only, and R=1.5 everywhere else (top), and Concentration versus time for the three middle observation points (Figure 4.6.), using SEAM3D-PUP and SEAM3D-RCT for case study (4.2.3.1)......................................................................................................................................................108
Figure 4.14. Mass removal by direct uptake and root sorption (top) and concentration (bottom) versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages for case study 4.3.1............................................................................................................................................................109
Figure 4.15. Conceptual model for case study 4.3.2, flow and transport with direct uptake in the middle ET area and root sorption all over the model with different values of RCF. .............................110
Figure 4.16. Concentration (top) and mass removal by direct uptake and root sorption (bottom) versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages for case study 4.3.2............................................................................................................................................................111
Figure 5.1. The expected effect of using a phytoremediation system on reducing DS concentration. 113 Figure 5.2. The conceptual model with the grid dimensions and boundary conditions..........................114 Figure 5.3. Source mass in the system vs. time (using SEAM3D-SSM and SEAM3D-PUP) under NA
conditions. ................................................................................................................................................115 Figure 5.4. ET rate for different stress periods. ..............................................................................................116 Figure 5.5. Initial conditions for the test models. ...........................................................................................118 Figure 5.6. Validation the results of SEAM3D-PUP by comparing the mass output of MT3DMS-SSM
versus PUP for a), solute mass in aquifer and b) solute mass removal for LET=0.5Lp and WET=300m. ..............................................................................................................................................119
Figure 5.7. Solute mass in the model domain for a) Different values of TSCF, and (b) The dynamically stable plume shows constant mass removal under NA conditions and oscillates around this value for TSCF = 0.0. (WET/Ws=3.0, LET=0.5Lp). ...........................................................................119
Figure 5.8. Groundwater hydraulic head profile showing the effect of phytoremediation.....................120
ix
Figure 5.9. Effect of ET width on solute mass removal for different values of TSCF: a) LET=Lp and b) LET=0.5Lp. ................................................................................................................................................122
Figure 5.10. Effect of TSCF on solute mass removal ET width values for: a) LET=Lp, and b) LET=0.5Lp.....................................................................................................................................................................123
Figure 5.11. Observation points for concentration profile. ..........................................................................124 Figure 5.12. Concentration profiles at distances = 500, and 1000 downstream the source for different
values of WET for a) LET=Lp and b) LET=0.5Lp, where TSCF=1.0................................................125 Figure 5.13. Concentration vs. distance at different observation points downstream the source at the
end of different stress periods for a) LET=Lp and b) LET=0.5Lp. ...................................................126 Figure 5.14. Concentration profiles for different TSCF values used to calculate the plume length at a
concentration = 1% of the source concentration for a) LET=Lp and b) LET=0.5Lp...................126 Figure 5.15. Comparison of the plume length under ET, (Lp*), to the plume length under natural
attenuation only, (Lp), for different ET dimensions (W/Ws) and TSCF values. ........................131 Figure 5.16. Concentration profiles at different times after the phytoremediation system starts for two
different LET. ............................................................................................................................................134 Figure 5.17. Reduction in plume length due to phytoremediation. .............................................................135 Figure 5.18. Effect of decay rate due to phytoremediation on the dissolved concentration. .................137 Figure 5.19. Calculating of mass-flux for the flow model of SEAM3D-PUP...........................................138 Figure 5.20. Distribution of right-face cell flow (out-flow), aqueous concentration and mass-flux at a
cross-section 500 m DS the source (WET/WS = 2.0). ......................................................................140 Figure 5.21. Mass-flux distribution at a cross-section 500 m DS the source for different TSCF values
for a) WET/WS =2.50), and b) WET/WS =3.0.....................................................................................141 Figure 5.22. Average Mass-flux results at different cross-sections downstream of the source for a)
LET=Lp and b) LET=0.5Lp for different values of TSCF, and WET=300. .....................................142 Figure 5.23. Average contaminant mass-flux at different cross-sections downstream the source for
LET=Lp and LET=0.5Lp, (WET=300, and TSCF=1.0). ....................................................................143 Figure 5.24. Average mass-flux reduction vs. (W/Ws) for different values of TSCF and LET. ..............143 Figure 5.25. Conceptual model for the study case 5-4. ..................................................................................145 Figure 5.26. Solute mass in the aquifer (or model domain) for different aquifer in-flux and ET lengths
(different out-flux) where the ET length starts at the source, TSCF=1.0. ..................................148 Figure 5.27. Solute mass in the aquifer (or model domain) for different aquifer in-flux and ET lengths
(different out-flux) where the ET length starts at the plume toe...................................................149 Figure 5.28. Comparison of solute mass in aquifer for different ET placement. .....................................149 Figure 5.29. Effect of out-flux, UET relative to in-flux, Uin on the solute mass removal. ........................150 Figure 5.30. Concentration profiles for aquifer in-flux (Qin=2.0 m3/d/cell) and different ET lengths
and locations. ...........................................................................................................................................152 Figure 5.31. Comparison for concentration profiles for different ET locations. .....................................153 Figure 5.32. Average solute mass-flux for different LET lengths and locations, Qin=200 m3/d. ..........155 Figure 5.33. Average reduction in solute mass-flux (with respect to the NA conditions) for different
LET lengths and locations, Qin=200 m3/d. ........................................................................................155 Figure 5.34. Comparison between mass-flux results for different phytoremediation system dimensions
and locations. ...........................................................................................................................................156 Figure 5.35. Effect of TSCF on the reduction of solute mass-flux (compared to the NA conditions) for
left and right locations of ET. ..............................................................................................................157 Figure 5.36. Effect of splitting the ET area into two halves on solute concentration and mass removal.
....................................................................................................................................................................159 Figure 5.37. Effect of splitting the ET area into two halves on solute mass-flux.....................................159 Figure 5.38. % Reduction in solute mass for different ET arrangements..................................................160
x
Figure 5.39. Concentration profiles at different time steps after the contaminant source is removed.162 Figure 5.40. Solute concentration profiles, source removed for LET=0.5Lp at left and right sides of the
plume footprint. ......................................................................................................................................163 Figure 5.41. Solute concentration profiles, source removed for LET=Lp, and comparison of the LET
location effect on concentration. .........................................................................................................164 Figure 5.42. Reduction in solute concentration (after the source is removed) for different LET lengths
and locations. ...........................................................................................................................................165 Figure 5.43. Solute mass in aquifer after removing the source, (a), and with a phytoremediation system
(b). ..............................................................................................................................................................165 Figure 5.44. Solute mass reduction due to applying a phytoremediation system where the contaminant
source is removed. ..................................................................................................................................166 Figure 5.45. Effect of WET on solute mass removal for different TSCF values for a) LET=Lp, and b)
LET=0.5Lp. ................................................................................................................................................168 Figure 5.46. Effect of the TSCF on solute mass removal for different values of (WET/Ws) for a) LET=Lp
and b) LET=0.5Lp.....................................................................................................................................169 Figure 5.47. Design charts for the ET width required to reduce the plume length to a certain design
value for different TSCF values for a) LET=Lp and b) LET=0.5Lp. ................................................170 Figure 5.48. Effect of TSCF on average contaminant mass-flux for LET=Lp and LET=0.5Lp. ...............171 Figure 5.49. Effect of WET/Ws on average contaminant mass-flux for a) LET=Lp and b) LET=0.5Lp. .172 Figure 5.50. Employing the design charts for a design problem..................................................................173 Figure 5.51. Estimating the phytoremediation system width for a given reduction in plume length. ..174 Figure 5.52. Estimating the value of TSCF for a given phytoremediation system width to reach a
certain reduction in plume length. .......................................................................................................175 Figure 6.1. Relationship of PCE in tree cores collected at the New Haven Site plotted versus the
groundwater concentration below each tree at (6 – 7.6 m). ...........................................................177 Figure 6.2. Relationship of PCE in tree cores collected at the New Haven Site plotted versus the soil
concentration 1.2 m below the surface near the base of the tree. .................................................177 Figure 6.3. The Langmuir nonlinear equilibrium isotherm. ..........................................................................182 Figure 6.4. SEAM3D-PUP results for ISO=2 for a) Solute mass removal, and b) solute concentration.
....................................................................................................................................................................185 Figure 6.5. Effect of TSCF using ISO-2 for a) Solute mass removal, and b) solute concentration for
initial source concentration = 10 mg/L, and N=0.75. ....................................................................186 Figure 6.6. Effect of starting concentration on mass removal using ISO-2 modeling option for a)
N=0.75 and different values of TSCF, and b) TSCF=1.0 and different values of N. .............188 Figure 6.7. Concentration (a), and solute mass removal (b) vs. time for different values of ISO-3
constant, K1 (Tc=8.0)...............................................................................................................................190 Figure 6.8. Effect of plant total concentration capacity, Tc on solute mass removal for ISO-3............192 Figure 6.9. Comparing the three different Isotherms. ...................................................................................192 Figure 6.10. Comparing SEAM3D-PUP alternative model with ISO=2, and N=1.0 and the linear
original code.............................................................................................................................................194 Figure 6.11. Mass-in aquifer (a), and solute mass removal (sinks) (b) for PCE with TSCF=0.7552 and
N = 0.787. ................................................................................................................................................195 Figure 6.12. Concentration profile for PCE. ...................................................................................................196 Figure 7.1 Linear and segmental ET packages. ...............................................................................................199 Figure A.1. Effect of ET width on solute mass removal, LET=Lp. ..............................................................220
xi
Figure A.2. Effect of ET width on solute mass removal, LET=0.5Lp..........................................................221 Figure A.3. Effect of TSCF on solute mass removal, LET=Lp......................................................................222 Figure A.4. Effect of TSCF on solute mass removal with different ET lengths. .....................................223 Figure A.5. Concentration profiles along the length of the plume for different values of TSCF at
different simulation times (5 yr, and 10 yr). .......................................................................................224 Figure A.6. Concentration vs. distance at different observation points downstream the source (with
exponential fitting in the bottom charts). ...........................................................................................224 Figure A.7. Concentration profiles for different TSCF values used to calculate the plume length at a
concentration = 1% of the source concentration for LET=Lp........................................................225 Figure A.8. Concentration profiles for different TSCF values used to calculate the plume length at a
concentration = 1% of the source concentration for LET=0.5Lp. .................................................226 Figure A.9 Average Mass-flux results at different cross-sections downstream the source for LET=Lp and
different values of WET and TSCF.......................................................................................................227 Figure A.10. Average Mass-flux results at different cross-sections downstream the source for
LET=0.5Lp and different values of WET and TSCF. ..........................................................................228 Figure A.11. Effect of the phytoremediation location and TSCF on solute mass removal....................229 Figure A.12. Concentration profiles for different aquifer in-flux (Qin=1.50 m3/d/cell) and ET lengths
....................................................................................................................................................................230 Figure A.13. Concentration profiles for different aquifer in-flux (Qin=1.05 m3/d/cell) and ET lengths
....................................................................................................................................................................231 Figure A.14. Effect of TSCF value on plume concentration for different ET locations........................232 Figure A.15. Average solute mass-flux for different LET lengths and locations, Qin=150 m3/d. .........233 Figure A.16. Average reduction in solute mass-flux (with respect to the NA conditions) for different
LET lengths and locations, Qin=150 m3/d. ........................................................................................234 Figure A.17. Comparison between mass-flux results for different phytoremediation system dimensions
and locations ............................................................................................................................................235 Figure A.18. Average solute mass-flux for different LET lengths and locations, Qin=105 m3/d. .........236 Figure A.19. Average reduction in solute mass-flux (with respect to the NA conditions) for different
LET lengths and locations, Qin=105 m3/d. ........................................................................................237 Figure A.20. Effect of inflow rate on solute mass-flux for different values of LET and ET locations..238 Figure A.21. Effect of in-flow rate on the reduction of solute mass-flux (compared to the NA
conditions) for different values of LET and ET locations................................................................239 Figure A.22. Effect of in-flow rate on the percentage reduction of solute mass-flux (compared to the
NA conditions) for different values of LET and ET locations........................................................240 Figure A.23. Effect of ET locations on the percentage reduction of solute mass-flux (compared to the
NA conditions) for different values of LET........................................................................................241 Figure A.24. Solute concentration profiles, source removed for LET=0.5Lp at left and right sides of the
plume footprint. ......................................................................................................................................242 Figure A.25. Solute concentration profiles, source removed for LET=Lp, and comparison of the LET
location effect on concentration. .........................................................................................................243 Figure A.26. Reduction in solute concentration (after the source is removed) for different LET lengths
and locations. ...........................................................................................................................................244 Figure A.27. Solute mass in aquifer after removing the source, (a), and with a phytoremediation system
(b). ..............................................................................................................................................................245 Figure A.28. Solute mass reduction due to applying a phytoremediation system where the contaminant
source is removed. ..................................................................................................................................245 Figure A.29. Solute mass-flux for different ET locations (up), and at downstream cross sections where
the contaminant source is removed.....................................................................................................246
xii
LIST OF TABLES
Table 1.1. Types of Phytoremediation Systems, (Miller, 1996 and Schnoor, 2002). .................................... 4 Table 2.1. Costs associated with various types of remediation methods (Wood, 2003)............................13 Table 2.2. Estimates of evapotranspiration rates by hybrid poplars .............................................................16 Table 2.3. Measured Transpiration Stream Concentration Factor (TSCF) and Root Concentration
Factor (RCF) for some typical contaminants and physical-chemical properties. .........................37 Table 2.4. Partition coefficient between octanol and water Kow for different chemicals. .........................39 Table 2.5. Root Concentration Factors (RCFs) of Pesticides and Related Compounds from Water into
Bode) Roots (Hordeum vulgare cv. Georgie) over a Period of 24 to 48 Hours and Calculated Quasiequilibrium Factors (αpt). ..............................................................................................................45
Table 2.6. Contaminant fate transport models comparison............................................................................58 Table 2.6. Contaminant fate transport models comparison, continued. ......................................................59 Table 2.7. Major Advantages and Disadvantages of the Phytoremediation Process. ................................63 Table 4.1. Comparison of concentration versus time from SEAM3D-PUP to both an exact Solution
and SEAM3D-SSM for the closed-system, single stress period model..........................................82 Table 4.2. Simulation results for mass removed by direct uptake and dissolved concentration versus
time using SEAM3D-PUP and five TSCF values for the closed-system model depicted in Figure 3.1. ...................................................................................................................................................82
Table 4.3. Comparison of concentration and mass removed through direct uptake versus time using SEAM3D-PUP to results using SEAM3D-SSM for the closed-system, two stress period model – case (3.1.2)...............................................................................................................................................82
Table 4.4. Simulation results for mass removed by direct uptake and dissolved concentration versus time using SEAM3D-PUP and five TSCF values for the closed-system model, four stress period model..............................................................................................................................................83
Table 4.5. Simulation results for mass removed by direct uptake for TSCF = 1.0 using SEAM3D-SSM and SEAM3D-PUP for the model shown in Figure 3.6. ..................................................................84
Table 4.6. Concentration results for the three observation points along ET zone for both SEAM3D-SSM and SEAM3D-PUP for TSCF = 1.0 – case study (4.1.3)........................................................85
Table 4.7. Simulation results for dissolved concentration versus time using SEAM3D-PUP and five TSCF values and compared to SEAM3D-SSM for the observation point (24, 50, 1) for the case study (4.1.3). ...............................................................................................................................................86
Table 4.8 Model parameters for the flow and transport with root sorption case study (4.2.1)................86 Table 4.9. SEAM3D-PUP and SEAM3D-RCT results for dissolved concentration at the observation
point (24, 50, 1) for the flow and transport with root sorption case study – f = 1.0...................87 Table 4.10. SEAM3D-PUP and SEAM3D-RCT results for mass removal at the observation point (24,
50, 1) for the flow and transport with root sorption case study (4.2.1) – f = 1.0.........................88 Table 4.11. SEAM3D-PUP and SEAM3D-RCT results for dissolved concentration at the observation
point (24, 50, 1) for the flow and transport with root sorption case study – f < 1.0, and f = 1.0.......................................................................................................................................................................89
Table 4.12. Model parameters for the flow and transport with root sorption case study (4.2.3). ...........89
xiii
Table 4.13. SEAM3D-PUP and SEAM3D-RCT results for dissolved concentration at the observation point (24, 50, 1) for the flow and transport with root sorption where 50% of the retardation factor is due to root sorption, and 50% is due to soil sorption – f = 1.0. .....................................90
Table 4.14. Concentration versus time for the three middle observation points (using SEAM3D-PUP and SEAM3D-RCT) for the model in Figure 4.6, with root sorption in ET area only for f = 1.0 (case study 4.2.3.1). ...................................................................................................................................91
Table 4.15. Results for mass removal by direct uptake and root sorption versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.1. ...................92
Table 4.16. Results for dissolved concentration versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.1. .................................................................93
Table 4.17. Dissolved concentration results for SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.2. ...................................................................................................94
Table 4.18. Results of mass removal by direct uptake and root sorption using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.2......................................95
Table 5.1. Summary of the variable model parameters and runs. Five values of TSCF (0.0, 0.25, 0.50,
0.75, and 1.0) were used in each case. .................................................................................................116 Table 5.2. Constant Model Parameters. ............................................................................................................117 Table 5.3. Observation cells (i, j, k)....................................................................................................................124 Table 5.4. Plume lengths at a concentration equals to 1% of the source concentration for ET length =
1000 m (approximately equals to the plume length). .......................................................................129 Table 5.5. Plume lengths at a concentration equals to 1% of the source concentration for ET length =
500 m (approximately half the plume length)....................................................................................130 Table 5.6. Phytoremediation area starts at the source (XET=0.0).................................................................146 Table 5.7. Phytoremediation area starts at the plume toe (XET is variable) ................................................146 Table 6.1. Toxic Effects on Hybrid Poplar (Populus deltoides × Populus nigra DN34) from Chlorinated
Aliphatic Compounds (Dietz and Schnoor 2001). ...........................................................................179 Table 6.2. Manual calculations of concentration and mass using manual calculations based on the
Freundlich model for the closed system test case.............................................................................184 Table 6.3. Mass, mass removal, and concentration results using the SEAM3D-PUP Freundlich model
for plant uptake for the closed system test case................................................................................184 Table 6.4. Mass removal for the closed-system test case using the SEAM3D-PUP Freundlich model for
plant uptake for different values of (N)..............................................................................................184 Table 6.5. Solute concentration in groundwater for the closed-system test case using the SEAM3D-
PUP Freundlich model for plant uptake for different values of (N). ...........................................184 Table 6.6. Manual calculations of concentration and mass using manual calculations based on the
Langmuir model for the closed system test case...............................................................................187 Table 6.7. Mass, mass removal, and concentration results using the SEAM3D-PUP Langmuir model
for plant uptake for the closed system test case................................................................................189 Table 6.8. Solute concentration at the end of the simulation and solute mass loss for different plant
total concentration, Tc. ...........................................................................................................................191 Table 8.1. Transpiration Stream Concentration Factors (TSCF) and Root Concentration Factors (RCF)
for selected ground-water contaminants. ...........................................................................................207 Table A.1. Average mass-flux at different cross-sections downstream the plume source for different
values of ET width. ................................................................................................................................247
xiv
C h a p t e r 1
Introduction
1-1 Background
Phytoremediation is the use of plants to remediate contamination in soil and groundwater. Plants
can be used to contain, remove, or degrade contaminants (USGS b, 2003). Figure 1.1 shows the
different processes taking part in phytoremediation (Keller, 2003). In the early 90’s, phytoremediation
emerged with promises of significant economies similar to those initially proposed for bioremediation.
Drawing upon geobotanical observations of metal accumulation by plants growing in areas
contaminated with metals such as nickel, the use of plants to extract and accumulate toxic heavy metals
was proposed. It was also proposed that toxic organic compounds might be degraded by the action of
microorganisms peculiar to the rhizosphere of plants (Environmental Cleanup, 2003).
PhotosynthesisO2
CO2 PhloemPhotosynthesis
+O2
H O2 Transpiration
Root Respiration
Exudation
Dark RespirationXylem
LignificationMetabolitesSequestration
Cometabolism
TranspirationH O, Nutrients, O2 2
Contaminant CO , H O, Cl2 2Mineralization
ContaminantUptake
H O, Nutrients2
CO , H O2 2O2
CO , H O2 2O2
O , CH , COOH, C H OH2 3 34
Figure 1.1. Schematic of phytoremediation processes.
1
The potential economic benefits of using plants for remediation are impressive. Plants are robust
and solar powered. Their root systems permeate soil and sediment environments with an extensive and
active membrane system. The soil near their roots has a microbial population that is orders of
magnitude greater than non-root soil. These benefits are provided with little or no maintenance
requirements. Furthermore, plant-based systems are welcomed by the public due to their superior
aesthetics and the societal and environmental benefits that their presence provides (Environmental
Cleanup, 2003).
The potential use of plants to remediate contaminated soil and groundwater has recently received a
great deal of interest. The science of phytoremediation arose from the study of heavy metal tolerance in
plants in the late 1980s. The discovery of hyperaccumulator plants, which contain levels of heavy
metals that would be highly toxic to other plants, prompted the idea of using certain plant species to
extract metals from the soil and, in the process, clean up soil for other less tolerant plants. Scientists
also found that certain plants could degrade organic contaminants by absorbing them from the soil and
metabolizing them into less harmful chemicals (Henry, 2000). More recently, engineers and scientists
have applied phreatophytes to the remediation of contaminated groundwater. Phreatophytes are plants
whose roots generally extend downward to the water table, which customarily feeds on the capillary
fringe.
Processes known to degrade/remove contaminates
A number of mechanisms (discussed in greater detail in Section 2-3) have been suggested to explain
why phreatophytes may be useful in clean-up contaminants, such as TCE, during phytoremediation:
• Phreatophyte roots may break down contaminants in soil through the effect of the enzyme
dehalogenase, root exudates that transforms or mineralizes contaminants (Schnoor, 1997).
• Phreatophytes and other plants may assist in the breakdown of contaminants in soil through
enhancement of microbial activity in the rhizosphere. Plant roots provide passive aeration,
serve as a nutrient source for microbes, and draw water to the surface (Lay, 1999).
• Plant tissues may accumulate contaminants. Poplars have demonstrated the ability to uptake
and store heavy metals in intracellular root spaces, and to translocate these compounds to
shoots and leaves (Hinchman, 1996).
2
• Phreatophytes may remove contaminants through metabolism, converting it all the way to
normal end points such as carbon dioxides and salts (Dietz and Schnoor, 2001).
• Poplars may provide a hydraulic control of aqueous contaminants, containing subsurface water
through uptake, thus decreasing the tendency of surface contaminants to move toward
groundwater. Poplars have been shown to transpire from 50 to 300 gallons of water per day
under some conditions (Chappell, 1997).
Contaminated water can be taken into the plant itself by direct uptake and stored in its structure. As
plants lose their leaves or die, the organic matter needs to be collected and transported to an
appropriate waste facility so that the contaminant is not reintroduced into the subsurface. Finally, plant
transpiration can help to provide hydraulic control of the site during the growing season. Transpiring
plants are known to create a depression in the water table; thus preventing contaminant migration by
forcing surrounding groundwater to flow towards the site. (Lay, 1999).
Table 1.1 lists the various applications of phytoremediation technologies. This list indicates that
phytoremediation is actually a broad class of remediation techniques that include many treatment
strategies. Obviously, the common thread through all of these techniques is the use of plants to treat a
contaminant problem. However, due to the diverse nature of chemical contamination and the diversity
of plants with the potential to treat them, remedial project managers must choose between wide
varieties of phytoremediation techniques to solve the problem at hand (Chappell, 1997).
Application of Phytoremediation to Groundwater Contaminants
For optimum effectiveness of phytoremediation systems, the various forms of phytoremediation
require different characteristics in the plants used. Poplar and cottonwood trees commonly are used
because they are fast-growing and have a wide geographic distribution. Examples of other types of
vegetation used in phytoremediation of surface soils include sunflower, Indian mustard, and grasses
(such as ryegrass and prairie grasses) (EPA, 2001). Figure 1.2 provides a cumulative overview of in situ
and ex situ treatment technologies selected for source control which phytoremediation represents (4)
sites in (in situ), and (4) sites in (ex situ) remedial sites, (EPA, 2001).
3
Table 1.1. Types of Phytoremediation Systems, (Miller, 1996 and Schnoor, 2002).
Treatment Method Mechanism Media Types of Plants
Rhizofiltration Uptake of metals in plant roots surface water and water pumped through troughs
Aquatic plants, (e.g., duckweed, pennywort), also Brassica, sunflower
Phytotransformation Plant uptake (sorption) and degradation of organics
surface water, groundwater
Trees and grasses
Plant-Assisted Bioremediation
Enhanced microbial degradation in the rhizosphere
soils, groundwater within the rhizosphere
Phytoextraction Uptake and concentration of metals via direct uptake into plant tissue with subsequent removal of the plants
soils Variety of natural and selected hyperaccumulators, e.g., Thalaspi, Alyssum, Brassica
Phytostabilization Root exudates cause metals to precipitate and become less bioavailable
soils, groundwater, mine tailings
Various plants with deep or fibrous root systems
Phytovolatilization Plant evapotranspirates selenium, mercury, and volatile organic compounds (VOC).
soils, groundwater Trees for VOCs in groundwater; Brassica, grasses, wetlands plants for Se, Hg in soil/sediments
Removal of organics from the air
Leaves take up volatile organics Air
Vegetative Caps Rainwater is evapotranspirated by plants to prevent leaching contaminants from disposal sites
Soils Trees such as poplar, plants (e.g., alfalfa) and grasses
Hydraulic control Plume capture/Phytotrans.
Removal of large volumes of water from aquifers by trees.
Groundwater Poplar, willow trees
4
Soil Vapor Extraction (213) 25%
Bioremediation (48) 6%
Solidification/Stabilization (48) 6%
Flushing (16) 2% ChemicalTreatment (12) 1%
Other (in situ) (27) 3%
In Situ Thermal Treatment (8)Multi-Phase Extraction (8)
Neutralization (4)Phytoremediation (4)
Vitrification (2)Electrical Separation (1)
Soil Vapor Extraction (9)Neutralization (8)
Soil Washing (8)Mechanical Soil Aeration (5)
Solvent Extraction (5)Open Burn/Open Detonation (3)
Phytoremediation (4)Vitrification (2)
Other (ex situ) (42)5%
Solidification/Stabilization (157)18%
Incineration (off-site)(104)12%
ChemicalTreatment (10)
1%
Thermal Desorption (69)
8%
Bioremediation (54)6%
Physical Separattion (20)2%
Incineration (on-site) (43)5%
Ex Situ Technologies (499) 58% In Situ Technologies (364) 42 %
Figure 1.2. Superfund Remedial Actions: Source Control Treatment Projects (FY 1982 -
2002).
Phytoremediation is a relatively new technology, for which there are only a few applications at
Superfund sites. Table 1.2 lists nine Superfund remedial action projects for which data on
phytoremediation are available. The technology is being applied to a variety of contaminants, including
halogenated VOCs, BTEX, chlorinated pesticides, radionuclides, and metals (EPA, 2001).
The most commonly used flora in phytoremediation projects are poplar trees, primarily because the
trees are fast- growing and can survive in a broad range of climates. In addition, poplar trees can draw
large amounts of water (relative to other plant species) as it passes through soil or directly from an
aquifer. This results in greater amounts of dissolved pollutants being drawn from contaminated media
and reduce the amount of water that may pass through soil or an aquifer, thereby reducing the amount
of contaminant flushed though or out of the soil or aquifer. In many cases, phytoremediation may have
a cost advantage over other treatment technologies because it relies on the use of the natural growth
processes of plants and often requires a relatively small investment in both capital and maintenance
costs (EPA, 2001).
5
Table 1.2. Superfund Remedial Actions.
Phytoremediation Projects FY 1982 - 1999, (EPA, 2001). Site Name (Operable Unit)
Contaminants (Target Cleanup Levels)
Media Type (a) Remediating Flora Status
Aberdeen Pesticide Dumps (OU5)
Benzenehexachloride (NR) Dieldrin (NR) Hexachlorohexane (NR)
Groundwater Hybrid Poplar Trees
Pre-design
Aberdeen Proving Grounds (Edewood Area, J-Field Soil OU)
1,1,2,2-Tetrachloroethene (NR) Trichloroethane (NR)
Soil and Groundwater
Hybrid Poplar Trees Magnolia Trees Silver Maple Trees
Operational
Boarhead Farm Cadmium (5 ug/L, Groundwater) Nickle (100 ug/L, Groundwater) Benzene (0.5 mg/kg, Soil) Trichloroethene (0.4 mg/kg, Soil)
Soil and Groundwater
NR Design
Bofors Nobel (OU1) Benzene (NR) Soil, Sludge, and Groundwater
NR Pre-design
Calhoun Park Area (OU1)
Benzene (NR) Toluene(NR) Ethylbenzene (NR) Xylene (NR)
Groundwater
Hybrid Poplar Tress
Operational
Idaho National Engineering Laboratory (USDOE, OU 21)
Chromium (NR) Cesium-137 (NR) Mercury (NR) Selenium (NR) Silver (NR) Zinc (NR)
Soil Prairie Cascade Willows Kochia Scoparia
Operational
Naval Surface Warfare Center, Dahlgren, Site 17
Mercury (<0.14 ug/L) Soil and Groundwater
Hybrid Poplar Trees Evergreen Trees
Pre-design
Naval Undersea Warfare Station (4 Areas, OU1)
1,1,1-Trichloroethane (NR) Groundwater Poplar Trees Operational
Tibbetts Road Trichloroethene (NR) Groundwater Poplar Trees Pre-design
NR - Not Reported (a) Treatments including both soil and groundwater are classified as source control treatments.
Phreatophytes
Phreatophytes are common in riparian habitats. The term literally means water-loving plants (such
as hybrid poplar trees). They are used to aid the breakdown of contaminants as well as control
contaminant transport. Plant roots in the soil increase the transfer of oxygen to the root zone. This, in
turn, promotes aerobic biodegradation of the contaminant in-situ. The rhizosphere (root zone)
encourages the growth of microbes in the soil that can use the contaminant as a carbon source
(WRRC, Arizona, 2003).
6
Phreatophytes have the ability to adapt to the desert conditions by developing extremely long root
systems to draw water from deep underground near the water table. Some roots have been recorded at
80 feet. Phytoremediation systems can also reduce recharge to the groundwater system due to the leaf
canopy (Hanson, 1991).
In phytoremediation systems, phreatophytes are used to control groundwater movement
(downgradient flux) by reducing recharge (the plants canopy reduces precipitations reaching the ground
surface, and thus reducing recharge), and increases evapotranspiration (ET). Phreatophyte-based
phytoremediation systems promote direct transpiration and reduce groundwater velocity and
contaminant flux, in some cases reversing the direction of groundwater flow. Removing groundwater
from an aquifer system creates a depression or a capture area that helps control the transport of
contaminants and remediate the groundwater.
Use of Trees (Phreatophytes) in Phytoremediation
Phytoremediation offers the potential for remediating groundwater and soil with the following
benefits (Quinn, 2000):
• Reasonably low installation cost, remediation within a suitable time frame, low operation and
maintenance costs, aesthetic value, low ecological impact, and public approval.
• In the last decade, hybrid poplars have been studied to determine their ability to remove or
destroy contaminants such as volatile organic compounds (VOCs).
• Other advantages of using poplars in certain phytoremediation systems include their fast
growth rates and their ability to use vast amounts of water.
• Poplars can achieve growth rates as high as 10 to 16 ft/yr (3 to 5 m/yr) (Chappell,1997).While
they can transpire tremendous amounts of water (Nyer and Gatliff,1996),the rate varies,
depending on climatic factors and tree density (Chappell 1997).Their ability to lower the water
table indicates that they have the potential to provide groundwater containment (Nyer and
Gatliff,1996; Compton et al.,1998; Newman et al.,1999).
Cunningham et al. (1997) described phytoremediation as the use of plants for “solar-driven
pumping and filtering systems” (though plants don’t actively transport TCE), with a root system that is
“exploratory, liquid phase extractors that can find, alter, and/or translocate elements and compounds”.
In the case of TCE contamination, hybrid poplars (Populus trichocarpa x Populus deltoides) and
Eastern cottonwoods (Populus deltoides) have received most of the attention. These are phreatophytic
7
species, meaning that their deep roots draw water from the water table. In addition, poplars and
cottonwoods have a fast growth rate, and have demonstrated an ability to take up TCE from both soil
and water (Lay, 1999).
Presently, there is no methodology or systematic analysis for the design of phytoremediation
systems for groundwater capture and contaminant control. Likewise, software tools for application to
existing sites are needed to determine the effectiveness of phytoremediation systems applied to real
cases of studies. A discussion of previous research will be presented in the literature review section.
This research more directly addresses the need for solute transport models that incorporate removal
and attenuation of contaminants from ground-water systems by plants.
1-2 Objectives
The goal of this research is to develop and validate a model that simulates the attenuating effects of
plants on aqueous-phase contaminants due to the specific mechanisms of plant uptake, root sorption,
and biodegradation. The model will be implemented in equations of 3D groundwater flow and solute
transport, which will be solved using a computer code.
Three specific research objectives are identified:
1- Develop a mathematical model for the removal and attenuation of aqueous-phase
contaminants by phreatophytes from groundwater systems. The model is implemented and
tested using the code SEAM3D (Sequential Electron Acceptor Model, 3D Transport).
2- Investigate the effect of different design scenarios for a poplar-based phytoremediation system
on hydraulic control, solute mass removal, and dynamic reduction in plume dimensions and
contaminant mass flux.
3- Extend the original SEAM3D-PUP code capabilities to include the simulation of plant uptake
for different mechanisms beside the linear model presented in the original code.
1-3 Organization of the Dissertation
The dissertation consists of eight chapters in addition to the bibliography and Vita. Chapter 1, the
executive summary identifies the research deficiency in the point of research, the research objectives,
approach. Chapter 2 is dedicated for the literature review for the research in using phytoremediation
for plume control with an emphasizes on plant uptake and root sorption models. Chapter 3 has the
model development stages including the conceptual and mathematical models for direct uptake and
root sorption and then the model implementation. Chapter 4 is assigned for model testing and
8
verification. The verification of the plant uptake/root sorption package included the following study
cases: Demonstration of plant uptake: closed system model – single stress period, closed system model
– multiple stress periods, and flow and transport with direct uptake. Demonstration of root sorption:
flow and transport with root sorption (f = 1.0), flow and transport with root sorption (f < 1.0), flow
and transport with root sorption (f = 1.0) and aquifer sorption, flow and transport with spatially-
variable root sorption (f = 1.0). Demonstration of direct uptake and root sorption: flow and transport
with plant uptake and root sorption (in the ET area only), flow and transport with spatially distributed
RCF and plant uptake. Chapter 5 includes the tested and verified model applications which involves
the following study cases: Effect of ET area (WET × LET), Effect of the source and ET flux rate, Effect
of dividing the et area into two halves, and Effect of removing the source on contaminant mass
removal, downstream plume concentration, and solute mass-flux. For each of the study cases, in
addition to representing and commenting on the results, a series of design charts are introduced to help
deciding on using a phytoremediation system to achieve certain remediation goals. Chapter 6 is
including the SEAM3D-PUP code modifications to count for different plant uptake mathematical
models and the effect of toxicity which may lead the plant to have a maximum capacity for solute
uptake. Chapter 7: Conclusions and recommendations for future research and Chapter 8: SEAM3D-
PUP input instructions for the plant uptake transport and root sorption package.
9
C h a p t e r 2
Literature Review
2.1 Introduction
In phytoremediation and plant/soil/water interaction models, there is no one single model that can
predict every process taking place. The scope of this thesis is the contaminant mass uptake of poplar
trees from the saturated zone and the effect of plant roots on the contaminant retardation and fixation
by sorption.
Environmental awareness has increased during the last 40 years, realizing that in the race for
development and wealth, society failed to protect natural resources of our planet. Disposal of industrial
wastes was done randomly, and without any regulations, and it was regarded as “a non-productive
function to be achieved at the least possible cost” (Cook, 1977). This attitude in dealing with industrial
wastes side by side with no governmental interference, led to massive contamination of groundwater
and soil at sites across the United States (Ward, 1999) and people witnessed many environmental
disasters such as the pollution of Lake Erie and Lake Ontario (International Joint Commission, 1970),
the discovery of toxic waste under the Love Canal community of New York, Figure 2.1, which became
a national symbol of pollution (Levine, 1982).
These incidents of widespread pollution gained considerable public attention and brought about
monumental changes in American society. The steps towards solving the problems of groundwater
pollution begin in the late 1960’s and early 1970’s. The Solid Waste Disposal Act of 1965 (SWDA) was
the first act that regulated waste on a national scale (Reed et al., 1992). The National Environmental
Policy Act (NEPA) was approved by the Congress in 1969 establishing a national policy for the
environment protection among American citizens.
10
Figure 2.1. Signs used in Love Canal community of New York.
In 1970, President Richard Nixon established the Environmental Protection Agency (EPA) as the
implementing arm of the NEPA. Other important legislation of the 1970's included the Clean Air Act
(CAA; 1970), the Federal Water Pollution Control Act (FWPCA, 1972), the Safe Drinking Water Act
(SDWA; 1974), and the Resource Conservation and Recovery Act (RCRA; 1976). As stated by Reed et
al. (1992), these acts and others passed by Congress provided for the “cradle to the grave” regulation of
hazardous waste. Congress later passed the Comprehensive Environmental Response, Compensation,
and Liability Act (CERCLA, commonly called Superfund; 1980) that enabled the federal government
to delegate the costs of remedial act ion to the parties responsible for hazardous waste violations.
Pressure to meet the new standards for environmental quality led whole industries to re-engineer their
fundamental processes an d products (Cunningham et al., 1997) and forced some companies out of
business (Cammarota, 1980).
The proper disposal of hazardous waste and the need to clean existing contaminated sites became a
productive function for many public and private institutions in light of the substantial fines and
penalties, which could be mandated by regulatory agencies. Government agencies and private industry
alike began a search for efficient, cost-effective technologies that could be used to remediate hazardous
waste sites, an initiative that remains to the present day.
11
Currently 300,000 to 400,000 hazardous waste sites in the United States require some future
remedial action (NRC, 1997). However, only the EPA recognizes an estimated 30,000 of these as
candidates for immediate treatment (Ensley, 2000). These sites may be polluted with inorganic
contaminants, organic contaminants, or more commonly mixtures of both. The remediation of all U.S.
hazardous waste sites in existence could cost as much as $1 trillion (NRC, 1997), but the estimated
expense for sites of immediate concern is much less.
The projected cost for remediation of areas containing mixtures of heavy metals and organic
pollutants is $35.4 billion over the next five years, whereas cleanup of sites contaminated with metals
only would cost $7.1 billion (Ensley, 2000). The high cost of hazardous waste cleanup is due in part to
the inefficiency and high cost of available technologies. Conventional remediation techniques are based
on civil and chemical engineering technologies including a wide variety of physical, thermal, and
chemical treatments, as well as manipulations to accelerate or reduce mass transport in the
contaminated matrix (Cunningham et al., 1997). Table 2.1 summarizes approximated costs, and
limitations of some of the remediation technologies. Table 2.1 shows that one of the primary driving
forces behind the search for alternative remediation technologies is high cost of conventional methods.
According to the NRC (1999), as cleanup at waste sites has proceeded, it has become evident that
despite the billions of dollars invested, conventional remediation technologies are inadequate. The lack
of commercially available technologies that can restore contaminated sites at reasonable cost has led to
increasing pressure to limit waste cleanups to sites that pose immediate risks to human health.
Bioremediation is a biological treatment method, which employs microbial populations in the
remediation of contaminated soils and groundwater. Certain vegetation can sustain an eutrophic soil
environment for the bioremediation of many priority pollutants. This method of bioremediation soils
and groundwater is popularly known as phytoremediation (phyto means green plants and trees), (Davis
et al., 1998).
12
Table 2.1. Costs associated with various types of remediation methods (Wood, 2003).
Type of Medium Remediation Method Range of remediation cost (in U.S $) soil = per cubic meter water = per 1000 gallons cleaned
In Situ Vitrification [1,3] 360 1,370 Soil Incineration [3] 200 1,500 Excavation and Landfill [3,5,7] 140 720 Soil Washing [1,3,4,6] 80 860 Soil Flushing [1,3] 50 270 Solidification and Stabilization [1] 40 200
Electrokinetic Systems [1,3] 30 290 Bioremediation [1] 10 310
Soil Bulk density = 1.3
Phytoremediation of Soil [3,5,6,7] <1 150 Activated Carbon [6] 120 210 Biosorption [6] 9 3,400 Reverse Osmosis [6] 3 3 Adsorption [6] 1 20 Membrane separation –filtration [6] 1 6 Rhizofiltration [3,7] <1 6 Ion Exchange [6] <1 2
Water
Chemical Precipitation [6] <1 2
1- Woods, 1997 2- Ensley, 2000 3- Glass, 2000 4- Dennis et al., 1994 5- Salt et al., 1995a 6- Black, 1995 7- Cunningham et al., 1997
Note: Reported costs are estimates from available data. All soils were assigned a bulk density of 1.3 for the purposes of comparison.
2.2 Phytoremediation
2.2.1 Applicability of Phytoremediation
Phytoremediation has been used to clean up metals, pesticides, solvents, explosives, crude oil,
polyaromatic hydrocarbons, and landfill leachates. Phytoremediation can be used in combination with
other cleanup approaches as a “finishing” or “polishing” step. Although some phytoremediation
applications are slower than mechanical methods and are limited to the depths that are within the reach
of the plant roots (EPA, 1998). Vegetation is aesthetically pleasing, improves the site’s appearance,
serves as wild-life habitat and site-health monitor, prevents erosion, traps sediments, acts as a sorption
and biodegradation sink for pollutants, and may be harvested for fuel and lumber. Phytoremediation,
as an in situ remediation strategy, is economically competitive and acceptable by regulators where
conditions are appropriate. Investigations to explore suitable plant species, transport processes, and
transformation processes are the backbone of this multibillion-dollar remediation technology. The U.S.
13
EPA is currently supporting several initiatives and research projects involving the use of vegetation for
bioremediation.
2.2.2 Hyperaccumulators
Some plants, which grow on metalliferous soils, have developed the ability to accumulate massive
amounts of the indigenous metals in their tissues without exhibiting symptoms of toxicity (Baker and
Brooks, 1989). Chaney (1983) was the first to suggest using these “hyperaccumulators” for the
phytoremediation of metal-polluted sites. However, hyperaccumulators were later believed to have
limited potential in this area because of their small size and slow growth, which limit the speed of metal
removal (Cunningham and Ow, 1996).
By definition, a hyperaccumulator must accumulate at least 1000 µg Ag-1 of Co, Cu, Cr, Pb, or Ni,
or 10,000 µg Ag –1 (i.e. 1%) of Mn or Zn in the dry matter (Reeves and Baker, 2000). Some plants
tolerate and accumulate high concentrations of metal in their tissue but not at the level required to be
called hyperaccumulators. These plants are often called moderate metal-accumulators, or just moderate
accumulators (Kumar et al., 1995). The lack of variable plant alternatives for phytoremediation seemed
to suppress the amount of phytoremediation research conducted between the mid 1980’s and the early
half of the 1990’s. The search for plants for phytoremediation centered on the Brassica family, to
which many hyperaccumulators belong (Cunningham et al., 1995). Through the work of various
researchers, particularly Kumar et al. (1995) and Dushenkov et al. (1995), several high-biomass, metal-
accumulating species were identified. Phytoremediation research gained momentum after the discovery
of these plants, and most of our understanding of this emerging technology has come from research
reports published since 1995.
2.2.3 Poplar Trees
Poplar trees are typically used in phytoremediation of organic pollutants because they are long
lasting (between 25 and 50 years), fast growing, hardy, and transpire large quantities of water. Poplar
trees can grow six to eight feet per year, reaching heights of 30 feet depending on species. For fast two
years of the tree life the expected transpiration could be 200 gallons per tree per year. Grown poplars
can uptake up to 100 liter per day of groundwater (Sutherson, 1997).
Phreatophytes can uptake water from the top of the saturated aquifer. As in a natural pump and
treat system, the tree root system of a phreatophyte will transpire water and draw down the water table
in the areas below the tree. However, a disadvantage of phytoremediation is that the roots must be able
14
to reach the contaminated groundwater for remediation, therefore, making phytoremediation an
unfeasible remedial technology for deep contaminated aquifers. Some companies such as
Treemediation® have patented systems to treat deep contaminated soil and groundwater. Table 2.2
lists recorded ET rates by poplar trees (Chappell, 1997).
Hybrid forms of the poplar tree have been utilized at sites with soil organic chemical contamination
of soil and groundwater. Most hybrid varieties are fast-growers, perennial, long-lived (25-50 years) and
tolerant of organic contamination (Schnoor et al., 1995). Poplar roots can extend towards the water
table and establish root mass that can potentially consume rather large quantities of water (Schnoor et
al., 1995). According to Edward Gatliff, founder of Applied Natural Sciences, poplar trees have the
ability to reach deep aquifers and pump 50 to 350 gallons per day (gpd) per tree (Matso, 1995). In
amenable soils and temperate conditions, hybrid poplars can grow 2 meters in the first growing season
and reach a height of 5 to 8 meters after 3 years (Schnoor et al., 1995).
In a study at the University of Iowa (Schnoor et al., 1995), exudates from hybrid poplar roots
contained 10 to 120 mg/L of dissolved organic carbon and 1 to 10 mg/L of acetic acid. An increased
amount of bioavailable substrates in the root zone is likely to support growth of larger populations, if
other factors are not limiting. Therefore, microbial activity could also be increased if poplars are
implemented into a treatment strategy.
Jordahl et al. (1997) reported the first evaluation of the effect of trees on microbial populations in
the rhizosphere. The rhizosphere soils of seven-year–old Imperial Carolina poplars were used to
enumerate five specific phenotypes. Total heterotrophs, denitrifiers, pseudomonads, BTX degraders,
and atrazine degraders were enumerated for three rhizosphere samples previously exposed to nitrate
and atrazine. The phenotypes were also enumerated in soil samples devoid of roots from an adjacent
cornfield.
15
Table 2.2. Estimates of evapotranspiration rates by hybrid poplars
Rate Source
100 to 200 L/day/tree (~26 to 53 gallon/day) for 5 year old trees Newman et al (1997) 100 L/day/tree for a 5 year old tree under optimal conditions Stomp et al. (1994) 13 gallons per day (estimated) when trees are calculated as low-flow Sheldon Nelson - Workshop on pumping wells
Phytoremediation of Organic Contaminants (1996)
1.6 to 10 gpd/tree (observed) sap flow rates for young hybrid poplars at the Aberdeen Proving grounds in Maryland
Compton (1997)
10 - 11 kg/tree/day (observed) in early summer for 1-2 year old Eastern cottonwoods growing in Texas
Greg Harvey (personal communication)
40 gallons per day (observed) for 5 year old trees in Utah in the summer Ari Ferro- Workshop on Phytoremediation of Organic Contaminants (1996)
In summary, the advantages of hybrid poplar trees as phytoremediation tools include:
• Extremely fast growing, hardy, and tolerant of high organics concentrations • Preformed root initials that allow rooting along the entire buried depth • Release of exudates that may stimulate active degrader populations of microbes • Direct uptake of organics and, in some cases, transformation to less toxic metabolites, (Aitchison et
al., 2000).
2.2.4 Phytoremediation Mechanisms
Phytoremediation is accomplished through different removal mechanisms depending on the nature
of the contamination and the type of vegetation. Some of these processes maybe predominant, and
others can happen in accompany with others. Schnoor (2002) summarized these mechanisms as:
a) Uptake and translocation:
The contaminants can be absorbed with groundwater to the plants through the roots membrane,
and move through the stems to the plants leaves, which is known by direct uptake. The
groundwater contaminants must be soluble to be uptaken. Translocation means moving from
one place to another, which means that the contaminants that were in the groundwater are now in
the plant structure. Some of the contaminants that can be degraded or removed using this process
are organics, lead, and BTEX.
b) Uptake and enzymatic phytotransformation:
In this process, groundwater contaminants are extracted from the soil, and then the nature of the
contaminants changes by plant metabolism. A biochemical reaction takes place inside the plants
to transform the contaminants to other harmful chemical forms. The phytotransformation takes
place in many steps (the contaminants can go through many biochemical reaction processes to be
16
transformed to the final phase). Some of the contaminants that can be degraded or removed using
this process are organics, lead, BTEX, and TCE.
c) Rhizosphere bioremediation:
Rhizosphere refers to the roots zone where bacteria are active. In this process, which requires that
the contaminants be close to the roots zone, biodegradation happens due to the interaction
between the contaminants and bacteria in the rhizosphere. In this process, contaminants can be
degraded without actually getting into the plant structure. Some of the contaminants that can be
degraded or removed using this process are nitrogen compounds, especially those used in soil
fertilizing.
d) Phytostabilization:
The traditional means by which metal toxicity is reduced at these metal-polluted sites is by in-
place inactivation, a remediation technique that employs the use of soil amendments to
immobilize or fix metals in soil. In this process, the plants roots are used to stabilize the
contaminants in soil, and prevent or reduce further movement. In this process, no chemical
reaction takes place with the contaminants, or they are getting into the plant structure.
Phytostabilization is best suited for metal contaminants where keeping them immobile would be
the best choice, because they don’t eventually degrade.
This technique is actually a modified version of the in-place inactivation method in which the
function of plants is secondary to the role of soil amendments. Unlike other phytoremediative
techniques, the goal of phytostabilization is not to remove metal contaminants from a site, but
rather to stabilize them and reduce the risk to human health and the environment. Plants chosen
for phytostabilization should be poor translocators of metal contaminants to aboveground plant
tissues that could be consumed by humans or animals. The lack of appreciable metals in shoot
tissue also eliminates the necessity of treating harvested shoot residue as hazardous waste (Flathman
and Lanza, 1998).
e) Phytoextraction:
In this process, plant species extract contaminants (especially heavy metal contaminants) and keep
them inside the plant structure (roots, stems and/or leaves). Those plants are known as
hyperaccumulators. Phytoextraction refers to extraction of heavy metals into plants. It is
important to extract the plants, and isolate them in some sort of a disposal facility, and thus
17
isolating those contaminants from soil. Some of the contaminants that can be removed using this
process are heavy metals.
The terms phytoremediation and phytoextraction are sometimes incorrectly used as synonyms, but
phytoremediation is a concept while phytoextraction is a specific cleanup technology. The
phytoextraction process involves the use of plants to facilitate the removal of metal contaminants
from a soil matrix (Kumar et al., 1995).
The time required for remediation is dependent on the type and extent of metal contamination, the
length of the growing season, and the efficiency of metal removal by plants, but normally ranges
from 1 to 20 years (Blaylock and Huang, 2000; Kumar et al., 1995). This technology is suitable for
the remediation of large areas of land that are contaminated at shallow depths with low to moderate
levels of metal- contaminants (Kumar et al., 1995; Wantanabe, 1997).
f) Rhizofiltration:
Metal pollutants in industrial-process water and in groundwater are most commonly removed by
precipitation or flocculation, followed by sedimentation and disposal of the resulting sludge
(Ensley, 2000). A promising alternative to this conventional clean-up method is rhizofiltration, a
phytoremediative technique designed for the removal of metals in aquatic environments. This
process takes place in plants roots when roots sorb contaminations by membrane phenomena
when liquids move from more concentrated to less concentrated solutions. As in phytoextraction
process, plants roots are collected when they accumulate too much contamination. This process is
used with radioactive contaminants.
Dushenkov and Kapulnik (2000) describe the characteristics of the ideal plant for rhizofiltration.
Plants should be able to accumulate and tolerate significant amounts of the target metals in
conjunction with easy handling, low maintenance cost, and a minimum of secondary waste
requiring disposal. It is also desirable for plants to produce significant amounts of root biomass or
root surface area.
g) Hydraulic control:
Hydraulic control refers to removing groundwater from an aquifer by transpiration, and thus
capturing contaminants through direct uptake. In the process, the groundwater flux down
gradient the contaminant area is reduced and the contaminant mass flux is consequently reduced
even if contaminant mass is not removed by the plant system. In this process, which uses deep-
18
rooted plants, the plants themselves function as pumping wells. This process is suitable for
volatile and semi-volatile organic compounds.
h) Phytovolatilization:
Phytovolatilization refers to removing contaminants through plants leaves following direct uptake
and phytrotranslation. Although most applicable to volatile organic compounds (VOCs), some
metal such as As, Hg, and Se may exist as gaseous species in environment. In recent years,
researchers have searched for naturally occurring or genetically modified plants that are capable of
absorbing elemental forms of these metals from the soil, biologically converting them to gaseous
species within the plant, and releasing them in to the atmosphere. Phytovolatilization is the most
controversial of all phytoremediation technologies. Mercury and Se are toxic (Suszcynsky and
Shann, 1995), and there is doubt about whether the volatilization of these elements into the
atmosphere is safe (Watanabe, 1997).
The phytovolatilization of Se and Hg into the atmosphere has several advantages. Volatile Se
compounds, such as dimethylselenide, are 1/600 to 1/500 as toxic as inorganic forms of Se found
in the soil (DeSouza et al., 2000). The volatilization of Se and Hg is also a permanent site solution,
because the inorganic forms of these elements are removed and the gaseous species are not likely to
be redeposited at or near the site (Heaton et al., 1998). Furthermore, sites that utilize this technology
may not require much management after the original planting. This remediation method has the
added benefits of minimal site disturbance, less erosion, and no need to dispose of contaminated
plant material (Heaton et al., 1998; Rugh et al., 2000). Heaton et al. (1998) suggest that the addition
of Hg(O) into the atmosphere would not contribute significantly to the atmospheric pool.
19
2.3 Models for Phytoremediation Processes
2.3.1 Plant Uptake
Water supplied to the plant by the root contributes to the overall water balance of the shoot.
Despite this important function of roots, relatively little is known about the processes that govern or
even regulate root water uptake. There is much evidence that the force driving water across roots is
usually provided by the tension (negative pressure) created by transpiration from the shoot and
extending to root xylem (Steudle, 1995; Tyree, 1997). Hence, the force driving water across the root
cylinder is usually a gradient in hydrostatic pressure, that is, water uptake requires an osmotic gradient.
This process of water and solute uptake is generally referred to as described in Section 2.2.4.
Water is not the only substance uptaken by the plant, but the soil and the groundwater is also
having solutes and minerals that are absorbed by the root and moves inside the plant stem into the
shoots (leaves and fruits). Figure 2.2 is showing the different solute concentrations in a controlled
volume of the plant, soil, and water. The solute concentration absorbed on the plants roots, CR
depends on the factor RCF, while as the solute concentration in the plants shoots, CT depends on the
factor TSCF.
CCR
CT
RCF
TSCF
Figure 2.2. Solute fate in plants.
20
Currently, a variety of models are available for predicting the uptake, translocation, and
elimination of organic contaminants by plants. These models can be applied to unsaturated, variably-
saturated, saturated flow, and range from simple deterministic risk assessment screening tools to
more complex models that consider physical, chemical, and biological processes in a mechanistic
manner (Fryer and Collins, 2003). The root water uptake is currently modeled in ecological,
hydrological, and atmospheric communities in different ways (Feddes et al., 2001):
• Local point-/field-scale ecological and hydrological modeling. • Considering the root system as a diffuse sink that penetrates each depth layer of soil uniformly. • Large-scale atmospheric modeling.
2.3.1.1 Local point-/field-scale models
The local point-/field-scale models are constructed around the plant root system. The models
investigate how plant roots work considering multiple vertical soil layers and by specifying details of the
root distribution and the soil hydraulic characteristics that determine water availability to roots. In
principle two alternative approaches can then be taken (Feddes 1981; Molz 1981).
The first plant-based approach is to consider the convergent radial flow of soil water toward and
into a representative individual root, taken to be a line or narrow-tube sink uniform along its length,
that is, of constant and definable thickness and absorptive properties. The root system as a whole
can then be described as a set of such individual roots, assumed to be regularly spaced in the soil at
definable distances that may vary within the soil profile. This microscopic approach that is
commonly used in ecological communities (Jackson et al. 2000b) casts the flow equation in
cylindrical coordinates and solves it for the distribution of soil water pressure heads, water contents,
and fluxes from the root outward. The problem with this approach is that often only steady-state
conditions are considered and that the required rather detailed plant information is often not
available.
2.3.1.2 Diffuse sink root models
The second more hydrologically oriented approach is to regard the root system as a diffuse sink that
penetrates each depth layer of soil uniformly, though not necessarily with a constant strength
throughout the root zone. Root water uptake can then be represented as a sink term that is added to
the vertical water flow equation through the soil. One has to realize, however, that one-dimensional
root system models may fail when lateral transport of water by subsurface or overland flow occurs. In
case of catchments with complex sloping terrain and groundwater tables, a vertical domain model has
21
to be coupled with either a process or a statistically based scheme that incorporates lateral water
transfer. This macroscopic way of solving the root water uptake problem is to combine the continuity
equation of water flow with a sink term representing water extraction by plant roots: Szq
t−
∂∂
−=∂∂θ ,
where θ is the volumetric water content [L3/L3], t is time [T], z is the vertical coordinate [L] taken
positively upward, q is the soil water flux [L/T] taken positively upward, and S is the sink term which is
representing the root water uptake rate [L3/L3 T-1].
When combining that equation with Darcy’s equation: ( ) ( )z
zhhKq∂+∂
−= , where K is the
hydraulic conductivity [L/T] and h is the soil water pressure head [L], results in Richards’ equation
(in one-dimension):
( ) ( ) ( )tzSzhhK
zthhC L ,1 −⎥
⎦
⎤⎢⎣
⎡⎟⎠⎞
⎜⎝⎛ +
∂∂
∂∂
−=∂∂ ………. (2.1)
where S(z,t) indicates that the sink term is a function of depth and time (Fayer, 2002). S can represent
the root water uptake rate.
The assumptions that led to the above equation are:
• Fluid is incompressible • Air phase is continuous • Air phase is at constant pressure • Flow is one-dimensional • Liquid water flow is isothermal • Vapor flow is negligible.
Examples of diffuse sink root models
SWMS_3D model
Hong et al. (2001) used two mathematical models to simulate phytoremediation effect on an MTBE
plume. Those models were: SWMS_3D and UNSAT-H. Both codes can be used for quantifying ET in
unsaturated zones. SWMS_3D is a computer program for simulating water and solute movement in
three-dimensional variably saturated media. The program numerically solves the Richards’ equation for
saturated-unsaturated water flow and the convection-dispersion equation for solute transport using
Galerkin-type linear finite element schemes. The flow equation incorporates a sink term to account for
22
water uptake by plant roots. The code allows simulation of time-varying root water and contaminant
uptake, surface evaporation, and infiltration.
SWMS_3D also provides a means for estimating actual transpiration as a fraction of potential
transpiration, based upon an experimentally determined “root-stress” curve provided by the user. This
root-stress curve comprises an attenuation factor, applied to potential transpiration that varies
depending upon the energy state (or head) of the water in the unsaturated zone (which can vary both
spatially and with time in the domain). A root (or sink) zone of any desired shape or size within the
domain could be assigned. Hence, roots concentrated near the ground surface or near the water table
can be simulated. Spatially varying local water uptake within the root mass may also be taken into
account by application of weighting factors (Simbnek, J., 1995).
The Richard equation ( qtrr
⋅∇+∂ θ ) in the θ-based form, where θ is used to represent the volumetric
water fraction and ( zKq +∇−= ψ )rr can be modified to the following form:
SKxhKK
xtA
izj
Aij
i
−⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛+
∂∂
∂∂
=∂∂θ ………. (2.2)
where:
θ = volumetric water content [L3/L3],
h = hydraulic head [L],
S = sink term [T-1],
xi = spatial coordinates [L], where (i=1,2,3),
t = time [T],
and are components of a dimensionless tensor KA representing the possible anisotropic nature of
the medium, and K is the unsaturated hydraulic conductivity function [L/T] given by:
AijK
( ) ( ) ( )zyxhKzyxKzyxhK rs ,,,,,,,, = ………. (2.3)
Where:
Kr = relative hydraulic conductivity
Ks = principal saturated hydraulic conductivity
23
The partial differential equation governing three-dimensional chemical transport during transient
water flow in a variably saturated rigid porous medium is taken as:
sswswi
i
jij
i
ScsCxcq
xcD
xats
tc
−++++∂∂
−⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
∂∂
=∂∂
+∂
∂ ργθγρμθμθρθ ………. (2.4)
where
c = solution concentration [ML-3],
s = adsorbed concentration,
qi = i-th component of the volumetric flux [L/T],
μw, and μs = first-order rate constants for solutes in the liquid and solid phases [T-1], respectively;
γw, and γs, = zero-order rate constants for the liquid [ML-3T-1] and solid [T-1] phases, respectively;
ρ = soil bulk density [ML-3],
S = sink term in the water flow equation
cs = concentration of the sink term [ML-3], and
Dij = dispersion coefficient tensor [L2T-1] (Simbnek, J., 1995).
SWAP Model
This previous modeling approach (Diffuse sink root models) is used in the models: Soil-Water-
Atmosphere-Plant, (SWAP) simulation model SWAP, and Unsaturated Soil Water and Heat Flow
Model , UNSAT-H. SWAP, The soil-Water-Atmosphere-Plant (SWAP) simulation model (Kroes et
al., 1999) is a transient, one dimensional model that uses soil physical properties, crop characteristics,
and weather hydrological data to estimate, on a daily basis, the components of the soil water balance
and the distribution of water within the profile, (Figure 2.3).
24
Precipitation
Soil evaporation
Surface runoff
Drainage/SubsurfaceInfiltiration
Drainage/SubsurfaceInfiltiration
Atmosphere
InterceptionTranspiration
UnsaturatedZone
SaturatedZone
Deep Groundwater
• Transport of: o Soil water o Soil heat o Solutes (salts, tracers)
• Influenced by: o Water repellency o Swelling and shrinkage o
Hysteresis
Plant
Figure 2.3. A schematic overview of the SWAP model system.
2.3.1.3 Large-scale atmospheric modeling
In general circulation models (GCMs) land surface parameterizations are often based on the
concept of a big leaf (Deardorff 1978), implying that the land represented in each grid element of the
model is homogeneously covered by a big leaf. However at the resolvable scale of GCMs land surfaces
are very heterogeneous. Avissar and Chen (1993) have therefore developed a set of prognostic
equations for momentum, heat, moisture, and other gaseous material quantifying mesoscale
circulations generated by landscape discontinuities and turbulent fluxes.
On the other hand various soil–vegetation–atmosphere transfer (SVAT) schemes have been
developed for use in GCMs and numerical weather prediction models. Their weakest component
however remains their link with the lower boundary. SVAT models face various difficulties, which
include (Kalma et al. 1999) comparable complexity between system components; scaling
incongruities between atmospheric, hydrological, and terrestrial components; and validation of
SVATs at appropriate timeand space scales. SVATs, which sometimes may be overparameterized,
use a variety of different methods to represent the relationship between roots, soil moisture and
transpiration. Moreover, SVAT parameters are generally highly variable in space and difficult to
measure. Because of all these reasons, it was not a surprise that the Project for Intercomparison of
Landsurface Parameterization Schemes showed that different SVATs/Land Surface Schemes (LSSs)
25
driven by the same meteorological forcing of air temperature, humidity, wind speed, incoming solar
radiation, longwave radiation, and rainfall can produce remarkably different surface energy and water
balances (Chen et al. 1997; Koster and Milly 1997; Pitman et al. 1999). The question in this context
was therefore raised: what is the role of roots?
2.3.1.4 Models for direct Transpiration
Evapotranspiration (ET) is a key process associated with plant uptake and plant-based
bioremediation (Davis et al., 1998). This phenomenon plays a significant role in sites with low
precipitation. ET prevents the percolation of precipitation into such contaminated site and draws up
the groundwater from the saturated zone. ET is the combination of two processes, evaporation from
the soil surface and transpiration from leaf surfaces of plants. ET depends on the plant species and
environmental factors such as temperature, wind velocity, and humidity. ET substantially influences
shallow water table levels of 2-5 m. When the water table is deeper (6-10 m), only deep-rooted and
drought resistant plants are able to send their roots down near the water table and pump up the water
(Pollock, 1994). This increases the net water flux and evapotranspiration by creating vertical water-
pressure gradients. This process clearly also transports the dissolved contaminants in the groundwater
to the unsaturated zone and into the roots. Sometimes, solar-driven transpiration translocates
contaminants into the stem of the plant. Boersma et a1. (1990) reported that accumulation of romacil
in plants increased in proportion to the transpiration rate.
MODFLOW
Evapotranspiration of groundwater may occur when the water table is close to the land surface or
when phreatophytes draw water from below the water table. Several groundwater flow models
incorporate losses from the saturated zone due to evapotranspiration, including the widely-used
Modular Groundwater Flow Code (MODFLOW). The Evapotranspiration Package of MODFLOW
requires the user to assign a maximum ET rate RETM to each cell from which ET may occur. The
maximum rate is used when the water table in a cell equals an assigned head value, normally equal to
the elevation of the land surface hs. No evapotranspiration occurs when the water table declines below
an assigned “extinction” depth (d). In between these two extremes the ET rate is assumed to be linear,
as shown in Figure 2.4 and 2.5 (Anderson and Woessner 1992 ; McDonald, and Harbaugh 1988).
The volumetric rate at which groundwater is removed by evapotranspiration is calculated in the
MODFLOW ET package as follows:
26
yxRQ ETMET ΔΔ= ………. (2.5)
where
ETMET QQ = For h > hs
0=ETQ For h > (hs – d)
( )[ ]d
dhhQQ sETMET
−−= For (hs – d) ≤ h ≤ hs
where h is the elevation of the water table calculated by the model, Figure 2.4. The extinction depth (d)
is normally 6 to 8 feet below the land surface but may be deeper if deep-rooted phreatophytes are
present. Thomas et al. (1989) set d equal to 12 feet beneath a playa in Nevada and 30 feet in the area
around the playa in which phreatophytes were growing. Danskin (1988) used 12 feet for the extinction
depth in Owens Valley in Southern California.
QETM
ETQ
hsh
d
MaximumEvapotranspiration
Slope = dETMQ______
0
Figure 2.4. Volumetric evapotranspiration, QET, as a function of head, h, in a cell where d is
the extinction depth, and hs is the ET surface elevation.
27
shh
d
(h - d )s
Figure 2.5. Representation of evapotranspiration in MODFLOW.
Matthews et al. (2002) investigated the effectiveness of phytoremediation through applying a PPS
on unconfined aquifer. The goal of their research was to develop a relationship between the plantation
area, and the capture of an aqueous contaminant plume. The model used constant head boundaries to
simulate the contribution of recharge to groundwater upgradient of the plantation area, Figure 2.6.
The following model assumptions were employed in Matthews et al. (2002):
1- The lower layers remain fully saturated. 2- Unconfined homogenous anisotropic sand aquifer. 3- Growing season period is 7 months (full effect of ET). 4- Hydraulic conductivity ranges were selected to match a site with predominately silty soil to
potentially sandy. 5- The authors used MODFLOW Recharge Package to simulate ET (Evapotranspiration) effect
by specifying a negative recharge rate within the footprint of the phytoremediation plantation. 6- The recharge rate outside the plantation area was set to 0.
Figure 2.6. Plan view of model grid (left) and cross section of model grid (right) used in
evaluating aquifer properties effect on phytoremediation effectiveness.
28
The main findings of Matthews et al. (2002) were:
1- The minimum plantation area for capture was found to be directly proportional to the ground
water flow rate (Q), and thus in direct proportion with (K, b, and I )
2- Nonlinear relationships were observed between the minimum phytoremediation plantation area
needed for capture and growing season duration, Figure 2.7. A location with a 9-month growing
season required 20% less plantation area than a location with the default 7-month growing
season, while a location with a 5-month growing season required 40% more of plantation area.
3- Higher aquifer anisotropy increases the phytoremediation area required to capture the plume,
Figure 2.8.
4- Wider plume width requires more phytoremediation area for capturing the plume, Figure 2.9.
5- The phytoremediation area needed to capture the specified contaminant plume was relatively
insensitive to specific yield and aquifer storativity. The minimum plantation area required for
capture varied <5% when the specific yield was varied between 0.01 and 0.25. Similarly, separate
simulations in which storativity was decreased by a factor of 3 and specific yield held constant
resulted in no significant change in the minimum plantation area required for capture. In
contrast, except in very permeable soils, the time required to develop the capture zone was
strongly dependent on aquifer storage.
6- Evapotranspiration fluxes through plantations, appropriately sized to contain the plume,
substantially exceeded the groundwater flux through the plume itself.
7- Phytoremediation may be impractical or not cost-effective in situations where the required
plantation area needed exceeds the amount of tillable land available.
29
Growing Season Duration, months3 6 9
Plan
tatio
n A
rea,
m2
4,000
8,000
12,000
16,000
12
Figure 2.7. Effect of growing season duration on minimum plantation area for capture.
0 50 100 150 200
Anisotropy Ratio
Sandy Soil
Silty Soil
Kh=0.0002 cm/s
Kh=0.0008 cm/s
Kh=0.002 cm/s
18,000
Plan
tatio
n A
rea,
m2
12,000
6,000
0
Figure 2.8. Effect of aquifer anisotropy on minimum plantation area for capture.
30
0
Plan
tatio
n A
rea,
m
2,000
4,000
6,000
2
0
Plume width, m30 60 90 120
Figure 2.9. Effect of plume width on minimum plantation area for capture.
The research of Matthews et al. (2002) did not employ the MODFLOW ET package. Therefore,
sensitivity of results to a number of input parameters was not considered:
1- Extinction depth (d): The model simulated ET by applying a negative recharge to the area of
plantation, which did not take into account the root extinction depth, Figure 2.10, McDonald
and Harbaugh, 1988.
2- ET rate (QET): The model used constant ET rate (by applying constant negative recharge)
3- Aquifer heterogeneity (K): The model was based on uniform values of hydraulic conductivity.
QET
M
ETQ
hsh
d
Max
imum
Evap
otra
nspi
ratio
n
Slop
e =
dETM
Q ____
__
0
shh
d
(h - d)s
Land surface elevation (SURF)
Figure 2.10. Effect of water table, and root depth on ET rate.
31
Potential problems in the design approach proposed by (Matthews et al., 2002) include:
1- Using negative recharge to simulate the plants uptake gave results indicating that vertical
anisotropy has a strong effect on phytoremediation area required to capture the plume. This
might be due to the fact that extracting water from the lower layers using the recharge package
in MODFLOW has to force water to come all the way from the bottom to the top, which is
somehow like a vertical one-directional movement where Kz plays an important role. Matthews
et al. (2002) mentioned, “Using the MODFLOW ET package, which incorporates a linear
dependence between ET rate and depth to the water table, resulted in varying groundwater
withdrawal rates from run to run, complicating comparisons”
2- The authors didn’t consider if extremes in transient site conditions such as seasonally varying
water table gradients, depth to groundwater, and seasonal or biological variations in ET rate
can be adequately represented in a steady-state model require or need additional research, but
the authors commented that “At a minimum, steady-state simulations are useful for screening
level design evaluations.”
3- Simulating the recharge using a constant head boundary made the water-table depth constant
without applying the phytoremediation simulation.
4- Using a constant ET rate while the ET rate is different by seasons.
5- The authors neglected the effects of vertical ground water flow gradients that might be induced
by precipitation recharge and/or local hydrogeologic conditions on phytoremediation
effectiveness.
6- This analysis did not explicitly consider the effects of local variations in the configuration or
degree of contamination of a ground water plume nor did we evaluate alternative plantation
layouts in addition to a simple rectangular design.
7- Just as optimal design for conventional ground water pump-and-treat systems is quite site
specific, we expect that plantation layout and plume configuration will have significant effects
on phytoremediation effectiveness. Because of the unique geometry of groundwater extraction
during phytoremediation, additional research is needed to evaluate this issue.
Matthews et al. (2002) research paper was an introductory effort to have an idea about designing
phytoremediation area required to capture a groundwater plume. The research is simplifying a lot of
32
parameters, and the design procedure should not be taken as is, but it needs more investigations for
other site-specific parameters.
MT3DMS
MT3DMS, Zheng (1999), is an update to the original MT3D, Zheng, (1990). MT3D stands for the
Modular 3-Dimensional Transport model, and MS denotes the Multi-Species structure for
accommodating add-on reaction packages. MT3DMS has a comprehensive set of options and
capabilities for simulating advection, dispersion/diffusion, and chemical reactions of contaminants in
groundwater flow systems under general hydrogeologic conditions. MT3DMS can be used to simulate
changes in concentrations of miscible contaminants in groundwater considering advection, dispersion,
diffusion, and some basic chemical reactions, with various types of boundary conditions and external
sources or sinks, Zheng (1999).
The partial differential equation describing the fate and transport of contaminants of species k in 3-
D, transient groundwater flow systems can be written as follows:
( ) ( ) ∑++∂∂
−⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
∂∂
=∂
∂nSS
ki
ij
k
iji
k
RCqCvxx
CDxt
C θθθ ………. (2.6)
Where
θ porosity of the subsurface medium, dimensionless
Ck dissolved concentration of species k, ML-3
t time, T
xi,j distance along the respective Cartesian coordinate axis, L
Dij hydrodynamic dispersion coefficient tensor, L2T-1
vi seepage or linear pore water velocity, LT-1; it is related to the specific discharge or Darcy flux
through the relationship, vi = qi θ
qs volumetric flow rate per unit volume of aquifer representing fluid sources (positive) and sinks
(negative), T-1
Csk concentration of the source or sink flux for species k, ML-3
ΣRn chemical reaction term, ML-3T-1
33
It is important to know that when using MT3DMS combined with MODFLOW ET package to
simulate solute plant uptake, the model does not take into consideration the factor TSCF and it
assumes 100% of solutes will be translocated from the saturated zone of groundwater table up to the
plant and hence to the atmosphere.
2.3.1.5 Equilibrium Models for Transpiration
Contaminant mass that enters the roots and does not accumulate there (as quantified by the RCF)
crosses the endodermis and enters the transpiration stream of the plant. Generally, plants translocate
water through their vascular bundles, which are mostly comprised of xylem and phloem. Studies have
revealed that solubility and volatility parameters are critical during iranslocation of organic compounds
in plants. Boersma et al. (1990) argued that the transfer of organic substances into a plant is primarily a
function of the lipophilicity (lipid loving potential) of the compounds.
The uptake advective flux is quantified by the transpiration stream concentration factor, TSCF,
which represents the ratio of the concentration of the compound in the transpiration stream within the
plant to the concentration of the compound in soil pore water (Briggs et al. 1982, Burken and Schnoor
1997, Trapp 1995).
C
CTSCF TS= ………. (2.7)
Where CTS is the concentration in the transpiration stream within the plant, and C is the solute
concentration in groundwater. The lowest possible value for TSCF is 0. Because passive uptake is
assumed for all xenobiotic compounds, the highest possible value for TCSF is 1.0 (Briggs et al. 1982,
Trapp 1995). TSCF has been shown to be independent of soil pore water concentration. Contaminant
flux into the transpiration stream can be calculated from transpirative water flux, soil pore water
concentration, and TSCF using
( )QCTSCFU = ………. (2.8)
Where U is the contaminant mass flux and Q is the transpirative water flux, provided mass is
eliminated from the plant shoots via metabolic degradation or volatilization out the leaves. If neither
metabolism nor volatilization occur, equation 2.8 does not apply and TSCF becomes a partition
coefficient which expresses equilibrium concentrations, (Schnoor 2002). Although the TSCF
contaminant flux concept is inherently steady-state, many previous efforts have applied the formula to
quantify uptake as part of a dynamic model (Behrendt et al. 1995, Burken and Schnoor 1997) which
34
implies that plants adjust their equilibrium to new environmental conditions fast enough such that this
approach is reasonably accurate.
TSCF values are also ultimately determined experimentally. The user can perform the experiments
directly for the compound and plant species of interest, rely on experimental data of others, or use
empirical equations based on curves fitted to experimental data that relate TSCF to chemical properties
such as the Kow for a specific plant. An example for barley roots was developed with O-
methylcarbamoyloximes and phenylureas by Briggs, et al 1982, Figure 2.11.
( ) ( )⎥⎦
⎤⎢⎣
⎡ −−×=
58.25.2logexp756.0
2owKTSCF ………. (2.9)
0.0-1.0 1.0 2.0 3.0 4.0 5.0
Tran
spir
atio
n St
ream
Con
cent
ratio
n Fa
ctor
, TSC
F 1.0
0.8
0.6
0.2
0.4
0.0
Log Kow
1-octanol/water partition coefficient
O-methylcarbamoyloximes
Figure 2.11. Relationship between the translocation of chemicals to barley shoots following
uptake by roots over 24 h (expressed as the Transpiration Stream Concentration Factor,
TSCF) and their 1-octanol/water partition coefficient (as log Kow); ο, O-
methylcarbamoyloximes; ×, substituted phenylureas.
35
Another example for poplar roots was developed with twelve organic compounds commonly found
at hazardous waste sites by Burken and Schnoor (1998):
( ) ( )⎥⎦
⎤⎢⎣
⎡ −−×=
44.278.1logexp784.0
2owKTSCF ………. (2.10)
It is recommended to use Briggs’ equation for herbal plants (experiments were done with the grass
barley), and Burken & Schnoors equation for woody plants (experiments were done on poplars),
(Trapp 2004).
For each of these equations, there is a maximum TSCF of about 0.8 in the moderately hydrophobic
range (Kow ≅ 100). At higher Kow values, TSCF decreases probably in part because compounds become
so hydrophobic that they sorb heavily to soil solids and root membranes. At lower Kow values, TSCF
decreases probably in part because compounds become so hydrophilic that they have trouble crossing
the lipid-rich root membranes (5a, 8). However, the TSCF concept is probably a simplification and
there may also be other factors at work (Briggs et al. 1982).
Experimental data in the literature, including that used to generate the equations for estimating both
RCF and TSCF discussed above, are mostly derived from plants grown in hydroponic solution in
laboratories. The accuracy of the values estimated by these equations varies due to the scatter in the
data used to derive them (Briggs et al. 1982, Burken and Schnoor 1998). In addition, one researcher
recently discovered a compound (1,4-dioxane) that significantly deviates from the TSCF equation’s
prediction (Aitchison et al. 2000). In this case, an unexpectedly high TSCF was observed for dioxane, a
fairly hydrophilic substance. The article suggests that the reason for this is that there are other
potential ways that hydrophilic substances can enter roots without having to bind and pass through the
lipid rich cell membranes (Aitchison et al. 2000).
RCF and TSCF values estimated from hydroponic experiments have been applied to estimate
uptake from soil water. This is generally reasonable because soil water is often in or close to
equilibrium with bulk soil concentrations, (Burken and Schnoor 1997, Trapp 1995), Table 2.3.
However, each application of this assumption should be evaluated separately (Burken and Schnoor
1997).
Caution should be exercised when selecting TSCF values for compounds that are known to degrade
metabolically in the transpiration stream of plants (e.g. atrazine), (Burken and Schnoor 1997). This is
36
because TCSF values are typically estimated by measuring mass emanating from plant leaves plus mass
accumulating in plant tissues – if significant degradation is occurring in tissues and that mass loss in not
being accounted for in the estimation of the TSCF value, the TSCF value, and hence any mass loss
from groundwater calculated using that TSCF value in (equation 2.8), may be erroneously low (Briggs
et al. 1982).
Table 2.3. Measured Transpiration Stream Concentration Factor (TSCF) and Root Concentration Factor (RCF) for some typical contaminants and physical-chemical properties.
+Log Kow
+ Solubility - --log C w sat
@25 °C, (mol/l)
+ Henry’s Constant K H , ,
@25 °C (dimensionless)
+ Vapor Pressure -log P o @25 °C
(atm)
Transpiration Stream Conc.
Factor (TSCF) †
(dimensionless)
Root Concentration Factor, RCF †
(L/kg)
Benzene 2.13 1.64 0.2250 0.90 0.82 1 Toluene 2.69 2.25 0.2760 1.42 0.81 3 Ethylbenzene 3.15 2.80 0.3240 1.90 0.80 2 m-Xylene 3.20 2.77 0.2520 1.98 0.78 11 TCE 2.33 2.04 0.4370 1.01 0.75 3 Aniline^ 0.90 0.41 2.2x10-5 2.89 0.32 420 Nitrobenzene 1.83 1.77 0.0025 a 3.68 0.82 3 Phenol* 1.45 0.20 >1.0x10 -5 3.59 0.48 12 Pentachlorophenol 5.04 4.27 1.5x10-4 6.75 0.04 30 Atrazine 2.69 3.81 1x10-7 9.40 0.57 8 1,2,4-Trichlorobenzene 4.25 3.65 0.1130 3.21 0.04 19 1,4-Dioxane -0.27 Miscible 2.0x10-4 0.05 0.72 <1 Methyl-tert-butyl ether 1.1 0.36 0.56 0.49 0.65 <1 TNT 1.90 3.36 - - 0.46 49 RDX 0.87 3.57 - - 0.16 1.3 HMX 0.19 4.77 - - 0.21 5.6 ^ pKa = 4.87, test conducted at pH 6.8 * pKa = 9.99, test conducted at pH 6.8 + Physical chemical properties (Schwarzenbach, et al., 1993) † Measured data from hydroponic studies with hybrid poplars (Burken and Schnoor, 1998; Dietz and Schnoor, 2001)
2.3.2 Root Sorption
The equilibrium partitioning between a hydrophobic phase (lipids, oils, etc.) and water is described
by the n-octanol-water partition coefficient Kow (L3/L3), which is a measure of the equilibrium
concentration of a compound of octanol and water that indicates the potential for partitioning into soil
organic matter (i.e., a high Kow indicates a compound which will preferentially partition into soil organic
matter rather than water).
CCK O
ow = ………. (2.11)
where CO is the equilibrium concentration of a substance in n-octanol (M/L3), and C is that in water
(M/L3). The Kow is used as a predictor for the partitioning between lipid phases in the environment and
water. Measured values are available for many compounds (Bedient 1994).
37
Kow is measured by mixing a chemical in an octanol and water solution the system is allowed to reach
equilibrium. The two phases will partition and a ratio of the chemicals concentration in the octanol
phase and water phase is taken. This ratio gives a relation of a chemicals accumulation in water. More
polar compounds will tend to have a low Kow. This is also a measurement of the hydrophobicity of an
organic. The more hydrophobic the more the contaminant will adsorb to soil and have a low solubility.
Kow is inversely related to the solubility of a compound in water. Kow is a dimensionless parameter and
usually ranges from 0.001 to about 100,000,000 and Log Kow is used in models to estimate plant and soil
invertebrate bioaccumulation factors. The Kow was first developed in the pharmaceutical industry and
has become a key parameter in studies of environmental fate of organic chemicals. Kow was found to be
related to water solubility, soil/sediment sorption coefficients, bioconcentration factors (BCF), (Leo et
al 1971, Bedient 1994).
The parameter Kow has been widely used to model organic compound uptake by plants because
octanot-water partitioning resembles the root tissues-soil water partitioning of many organic
compounds. If an organic compound has a log Kow < 1, it is highly water soluble. Such contaminants
are also mobile in plant xylem and phloem. Plants seldom accumulate these compounds beyond the
rare at which they are passively taken up into the transpiration stream. Nitroguanidine is one such
example. These contaminants usually are not targets for bioremediation studies using plants, (Schnoor
et al 1995).
Contaminants with log Kow between approximately l and 4 are generally xylem mobile and immobile
in phloem. Compounds of this type are expected to be good targets for bioremediation. Many of the
priority pollutants listed by U.S. EPA fall in this category with log Kow between 1 and 4. Compounds
with log Kow greater than 4 are plant xylem and phloem immobile. They tend adsorb onto root surfaces
and not be translocated to shoots of the plant. Most of the polyaromatic hydrocarbons compounds
(PAHs) fall in this category (ITRC 2001). Table 2.4 lists the measured in-lab values for Kow for different
chemicals, (Bedient et al. 1994, Gallagher 1998,Trapp 2004).
38
Table 2.4. Partition coefficient between octanol and water Kow for different chemicals. Octanol-water partition coeff. (at 20 °C) Chemical
Kow log Kow MTBE 13.8 1.14 Benzene 135 2.13 Toluene 490 2.69 o-Xylene 589 2.77 p-Xylene 1413 3.15 Ethyl benzene 1413 3.15 m-Xylene 1585 3.20 Terbutylazine 1622 3.21 Parathion 6457 3.81 Anthracene 28184 4.45 DDT 954993 5.98 Benzo(a)pyrene 1 348 963 6.13
2.3.2.1 Equilibrium Concentrations
If a substance is dissolved until it is a soluble in two adjacent, non-mixable phases (such as
groundwater and roots), the ratio of concentration in these two phases will have a certain value. The
calculation of equilibrium partition coefficients allows the estimation of the partition tendency of a
chemical. This concept has been quite successful for the estimation of chemicals’ fate. Together with
diffusion and advection processes, it is the basis of almost all exposure models, (Trapp, 1998).
A large amount of research has been devoted to understanding the parameters involved in
sorption/desorption of contaminants to soils and/or sediments. In one widely cited study, researchers
derived an equation for the partition coefficient for hydrophobic solutes between sediment organic
carbon and the aqueous phase (Karickhoff, Brown and Scott, 1979). The organic carbon partition
coefficient (Koc) is a measure of the tendency for organics to sorb onto the soil (or sediment) and is
defined as the ratio of the amount (mass) of a chemical sorbed per unit mass of organic carbon in the
soil or sediment to the concentration of the chemical in the soil (or sediment) solution at equilibrium,
CCK a
oc = , in which Ca is the concentration adsorbed (mass chemical adsorbed / mass organic carbon)
and C is the concentration in water (mg chemical / L H2O), (Fetter 1999). The performance of
bioremediation diminishes as Koc increases due to the lower bioavailability of contaminants strongly
sorbed to natural organic matter, (Looney, 2000).
39
In the study of (Karickhoff, Brown and Scott, 1979), the partition coefficient was related in two
separate equations between the octanol-water coefficient Kow and aqueous solubility (S),
21.0log00.1log −= owoc KK , and
………. (2.12)
44.0log54.0log +−= SKoc
where S represents the aqueous solubility expressed as mole fraction. Koc was then related by definition
to the partition coefficient (Kd) between the total sediment and the aqueous phase,
oc
doc f
KK = ………. (2.13)
where foc represents the mass fraction of organic carbon in the soil. The partition coefficient, Kd
describes the equilibrium distribution of a chemical between solids and groundwater. This is usually
described as a sorption isotherm between the concentration of the chemical sorbed onto the soil and
the concentration remaining in solution at equilibrium, (ASTM E1943).
CCK s
d = ………. (2.14)
where Kd is the distribution coefficient (L3/M), Cs is the sorbed concentration (M/M of soil), and C is
the dissolved concentration (M/L3 of the groundwater).
By employing such equations, researchers could then estimate the equilibrium concentrations of a
broad range of solutes based upon the Kow and/or solubility. Furthermore, the study found that the
linear partition coefficients were relatively independent of sediment solute concentrations and ionic
strength of the aqueous suspensions (Karickhoff et al. 1979).
Retardation resulted from sorption is defined as the process by which the movement of a reactive
chemical through an aquifer or geological unit is slowed or impeded due to sorption, (ITRC, 2002). It
is important for in situ bioremediation systems because retardation is a numeric value used to describe
the attenuation of a plume to sorption. If a contaminant is heavily retarded, it may not be available for
in situ bioremediation to occur, (ITRC, 2002).
40
Retardation is expressed in terms of the retardation coefficient, R:
e
db
nKR ×
+=ρ1 ………. (2.15)
where ρb is the bulk density of the soil matrix (M/L3), Kd is the partition coefficient, and ne is the
effective porosity (L3/L3). The retardation factor represents the transport velocity of the chemical
relative to the velocity of groundwater flow. The transport velocity of the chemical in groundwater, vc,
can be derived from R by: Rvvc = , where v is the groundwater velocity, and vc is the velocity of
chemical in groundwater, (ASTM, E1943-98).
These values are important to in situ bioremediation design to determine the degree of
contamination. Determining dissolution, retardation, and velocity help evaluate the feasibility or
enhancement of in situ bioremediation. Comparison of conservative tracers (bromide, chloride) with
contaminant movement can assist in velocity determinations, (ITRC, 2002)
2.3.2.2 Equilibrium Plant uptake Models
Equilibrium means that the whole chemical mass in the system is distributed between
compartments according to the equilibrium partition coefficient.
For the steady-state modeling process the input = output, 0=dtdm , and for the dynamic modeling
process (transient), or OutputInputdtdm
−= .
Trapp, (1995), summarized the equilibrium modeling processes by introducing the concepts of
connected compartments, Figure 2.12. Each compartment represents one media, i.e. air, water, soil,
……. Etc. The main levels of modeling are:
1- Level 1: Equilibrium, no reactions, closed system
2- Level 2: Equilibrium, open system, reactions, steady-state.
3- Level 3: Non-equilibrium, open system, reactions, steady-state.
4- Level 4: Non-equilibrium, open system, reactions, non-steady-state.
41
h
b
h
b
input
output
h
b
input
output
input
output
(1)
(2)
(3)
Comp. 1 Comp. 2 Comp. n
Figure 2.12. Equilibrium modeling levels.
The whole compartments are forming what’s Trapp, (1998) called an (Environmental Segment),
where each and every solute concentration can have different phases (similar to soil/water/air block)
and the concentration in each phase can be calculated based on different equations, and then they are
related together through the equilibrium concept.
For example, Trapp, (1998), in the model CemoS, used Richards’s equation to calculate pressure
head in partially saturated soil, and then used the equations of Dispersion/advection for first order
degradation:
( ) ( ) ( tDtutx
DtA
mtxC λ
π−⎥
⎦
⎤⎢⎣
⎡ −−= exp
4exp
4,
2
) ………. (2.16)
Where:
C(x,t) Concentration of the chemical substance at x and t, (M/L3).
x Coordinate in the flow direction, (L).
t Time after release, (T).
m Mass of the chemical substance released, (M).
42
A Cross-sectional area, (L2)
D Dispersion coefficient (L2/T)
u Flow velocity in the direction of flow (x-direction), (L/T)
λ First order reaction rate constant, (T-1).
Then the concentration in the groundwater phase is related to the concentration in the plant roots
by the factor, RCF, where CCRCF R= , where CR is the concentration sorbed in and on the roots, and
C in the concentration in soil pore water.
2.3.2.3 Sorption/desorption Kinetics
Researchers have argued that models based solely upon equilibrium do not adequately describe the
sorption/desorption processes of fluctuating systems such as frequently flooded topsoils. One model
describes the kinetics of sorption/desorption based upon not only Kow and organic carbon content but
also solution diffusivity, soil density, and soil porosity (Wu and Gschwend. l986). These researchers
found that the rate of hydrophobic compound desorption decreases with increasing Kow, organic
carbon, and aggregate size, and increases with water flaw. Other researchers found that the
sorption/desorption kinetics of aged organic compounds were temperature dependant
(Comelissenetal.,1997). Colder systems, it was found tended to retain sorbed contaminants longer than
warmer systems. Hence, although the temperature dependence of hydrophobic contaminant
biodegradation is often attributed to the temperature dependence of biological activity itself (Ghadiri.
Rose and Connell 1995), hydrophobic contaminants are also less likely to be bioavailable under cooler
conditions.
PCB’s in particular have engendered a spate of recent sorption kinetics research. It has been
reported, for example, that PCB’s tend to deserts in a two-phase model, whereby PCB’s deserts from
sediments first relatively quickly, then slowly over an extended period (Ghosh et al. 1999).
The desorption rate constants for the labile pool were found to be two orders of magnitude higher
than the rate commands for the slowly describing deal. Both pools, however, were shown to desarb
mare slowly with increasing overall chlorination. decreasing nation chlorination, and decreasing
temperance. This study was in agreement with an earlier study, wherein PCB contaminated sails were
submerged into water and the relative PCB desorption rates were measured (Girvin et al , 1997). In the
earlier study, the labile fraction was found to consist of 80-90 % of the total PCB concentration, and
43
most of this fraction desorbed within 48 hours of contact with water. Although this study
demonstrated that PCB’s were able to reach equilibrium in a matter of hours or days. It should he
noted that the organic content of the soils studied was relatively low (<0 2 %) and likely had a large
impact on the desorption kinetics.
2.3.2.4 Root Concentration Factor, RCF
In early studies, Lichtenstein (1959) found that lindane in soil was taken up by root crops (e.g.,
carrots and potatoes) more readily from light mineral soils than from a muck soil. Similarly, Walker
(1972) showed that the concentrations of atrazine in shoots of wheat plants growing in 12 different
soils were inversely proportional to soil-organic-matter (SOM) contents. In a more specific study on
the effect of soil type on crop uptake, Harris and Sans (1967) compared the levels of dieldrin
accumulated by carrots, radishes, and other root crops from three well-characterized contaminated
field plots in relation to the soil pesticide levels; the three soil types studied—a sandy soil, a clay loam,
and a muck soil-differed widely in SOM content (1.4 to 66.5%) and other soil constituents. Plant
dieldrin concentrations were much lower for crops from the muck soil than from sandy and clay soils;
by contrast, soil dieldrin concentrations were considerably higher in the muck soil than in the two other
soils.
For plant uptake of contaminants from soil-free nutrient solutions, Briggs et al. (1982) measured the
uptake by barley roots of two series of organic compounds, O-methylcarbamoyloximes and substituted
ureas, which vary widely in lipophilicity. They concluded that the root uptake of both types of
compounds approached the equilibrium values in a relatively short time (24 to 48 h).
Rhizosphere bioremediation and rhizofiltration require contaminants to be associated on or near the
roots. Briggs, et al. (1982) defined the Root Concentration Factor (RCF) as the ratio of organic
chemical sorbed on the root (mg/kg of fresh root tissue) to that in hydroponic solution (mg/L), or
CCRCF R
*
= , where CR* is the concentration sorbed in and on the roots, and C in the concentration in
soil pore water. It also typically includes partitioning to water in the root interiors, but this portion is
negligible except for hydrophilic substances, thus, the slope of a linear sorption isotherm is a measure
of the RCF and has units of L/kg (mL/g dry roots). Table 2.5 summarize the measured values of log
Kow, and RCF of Briggs, et al. (1982) quoted from (Cary, 2001).
44
Table 2.5. Root Concentration Factors (RCFs) of Pesticides and Related Compounds from Water into Bode) Roots (Hordeum vulgare cv. Georgie) over a Period of 24 to 48 Hours and Calculated Quasiequilibrium Factors (αpt).
Compound log Kow RCF αpt
O-Methylcarbamoyloximes
Aldoxycarb -0.57 0.66 0.74
Oxamyl -0.47 0.91 1.02
Acetone O-methylcarbamoyloxime -0.13 0.95 1.06
Aldicarb 1.08 0.94 0.90
Benzaldehyde O-methylcarbamoyloxime 1.49 1.48 1.19
4-Chlorobenzaldehyde O-methylcarbamoyloxime 2.27 2.80 0.98
3,4-Dichlorobeozaldehyde O-methylcarbamoyloxime 2.89 5.61 0.64
3-Phenylbenzaldehyde O-methylcarbamoyloxime 3.12 8.72 0.61
3-(3,4-Dichlorophenoxy)benzaldehyde O-methylcarbamoyloxime 4b 81.1 21
Substituted ureas
3-Methylphenylurea -0.12 0.73 0.82
Phenylurea 80 1.20 1.25
4-Fluorophcnllurea 1.04 1.10 1.06
3.(Methylthio)phenylurea 1.97 0.94 0.72
4.Chlorophenylurea 1.80 2.00 1.28
4.Bromophenylurea 1.98 3.17 1.63
3,4-Dichlorophenylurea 2.64 5.86 1.09
4-Phenoxyphenylurea 2.80 7.08 0.97
4-(4-Bromophenoxy)phenylurea 3.7 34.9 0.68
The root concentration factors (RCFs), increased monotonically, but not proportionally, with the Kow
values of the compounds. Similar empirical correlations for contaminants in plant roots and leaves
were also observed (Trapp 1995). In view of the influences of soil type and contaminant identity on
plant uptake, the plant contaminant levels is related to physico-chemical properties of the contaminants
and to the properties and compositions of plants and soils.
Most current models for plant uptake of contaminants from soil, water, or air are formulated on a
differential mass-balance basis in terms of the rates of contaminant interface transfer, plant growth and
transpiration, and contaminant metabolism, along with some estimated transfer coefficients (Riederer,
1990; Trapp et al., 1990; Paterson et al., 1994; Trapp and Matthies, 1995; Tam et al., 1996). Although
these models are intended primarily for delineating the rates of contaminant uptake by plants (or their
specific parts) with time from given external source(s), the model calculations are very sensitive to the
45
accuracy of assumed contaminant interface-transfer rates and coefficients. Alternatively, equilibrium
models have been utilized in some studies to assess contaminant levels in plants (or their parts) after
their exposure to chemicals in water over a certain period of time (Briggs et al., 1982; Trapp, 1995).
However, the actual state of a contaminant in plants may or may not be at equilibrium with the external
source, (Chiou, 2003).
A quasi-equilibrium partition model has recently been developed by Chiou et al. (2001) to account
for the passive plant uptake of contaminants from their external sources in soil or water. The model
takes explicit account of the plant contaminant level in relation to the source level and plant
composition. Moreover, the model contains both equilibrium and kinetic features and sets the upper
(equilibrium) limit for the level of a contaminant in a plant with respect to the external-source level,
against which the actual approach to equilibrium of the contaminant in the plant at the time of analysis
can then be estimated. Although in the initial model testing by Chiou et al. (2001) the partition
coefficients of contaminants with certain plant components have had to be estimated, the observed
consistency of the plant-uptake data with the conceived model parameters is stimulating to warrant
further investigation. The essential features of the model are presented below.
Organic chemicals with log Kow values greater than 3.0 are strongly sorbed to roots. Table 2.3
provides a number of organic chemicals, their physical chemical properties, and measured RCF values
on hybrid poplar roots. Of these, pentachlorophenol and 1,2,4-trichlorobenzene are strongly sorbed to
root tissues based on hydrophobic partitioning, (Schnoor, 2002). However, contaminants can be
immediately transformed at the root surface by extracellular enzymes or by membrane-bound enzymes.
Two exceptions to the governing rule of hydrophobic interactions at the root-water interface are
aniline and phenol (Table 2.3), (Schnoor, 2002). These compounds bind irreversibly to the root
(especially aniline) and are chemically transformed. They are not appreciably desorbed because they are
covalently bound as metabolic products in plant tissue (Lang, 1998; Hughes, et al., 1997). Other
examples include the reduction of nitroaromatic explosive compounds such as 2,4,6-trinitrotoluene,
(Hughes, et al., 1997 and Thompson et al. 1998). Benzotrizoles in aircraft deicing fluids appear to be
taken up and incorporated into the lignin fraction of the plant (Castro, et al., 2001; Castro, et al., 2000).
RCF values are determined by experiments. Empirical regression equations indicate RCF increases
with the octanol-water partition coefficient, Kow. An example for barley roots was developed with O-
methylcarbamoyloximes and phenylureas by Briggs, et al. (1982) showed that the greater the
hydrophobicity of the organic chemical, the greater was the tendency for sorption.
46
( ) 52.1log77.082.0log −=− owKRCF , or
………. (2.17)
( ) 77.00302.082.0 owKRCF +=
Figure 2.13 represents the measured, and fitted curve for RCF in terms of log Kow, (Briggs et al., 1982).
Burken and Schnoor (1998) published a similar relationship for twelve organic contaminants
typically found at waste sites with hybrid poplar roots grown hydroponically.
( ) 57.1log65.00.3log −=− owKRCF , or
………. (2.18)
( ) 65.0027.00.3 owKRCF +=
Also Trapp, S. in 2004, presented a similar equation to calculate the partition coefficient of roots to
external solution, KRW (units = mass per volume/mass per volume), which describes the equilibrium
partitioning between root concentration CR (mg/kg of fresh root weight) and water concentration, CW
(mg/L). The partitioning occurs into the water, the lipid and the gas phase of the root according to the
equation:
( ) ( ) AWAW
RbowRRRW KrootPKaLWK ++=
ρρ ………. (2.19)
Where W and L are water and lipid content of the plant root, “b” is a correction exponent for
differences between plant lipids and octanol, for roots is 0.77, 22.11tan
==oloc
a ρ . ρR is the density
of the fresh root, and ρW is the density of the external solution. Partitioning into the gas phase of the
root, PA(root), is usually negligible.
47
Log Kow
-1.0 0.0 1.0 2.0 3.0 4.0 5.0
1.0
10
100
Root
Con
cent
ratio
n Fa
ctor
, RC
F
1-octanol/water partition coefficient
O-methylcarbamoyloximes
Figure 2.13. Relationship between the uptake of chemicals by plant roots (expressed as the
Root Concentration Factor, RCF) from nutrient solution at 24 h and their 1-octanol/water
partition coefficient (as log Kow) for O-methylcarbamoyloximes and substituted phenylureas.
2.3.3 Rhizosphere Biodegradation
The plant root zone (rhizosphere) is providing a rich natural environment for bacteria to biologically
remediate the contaminants. Simulating the effect of plant roots on contaminants biodegradation is
no different from other known biodegradation simulating software packages. SEAM3D, for instance,
is having a biodegradation package which can be used to simulate the plant root effect.
2.4 Research on Phytoremediation
Phytoremediation of organic contaminants has generally focused on three classes of compounds:
chlorinated solvents, explosives and petroleum hydrocarbons (PHCs). Banks et al. (1997) and E. N.
Drake (1997) have conducted pioneering research into the phytoremediation of petroleum
hydrocarbons. Jerald Schnoor at the University of Iowa has done extensive studies on the uptake of
chlorinated and explosives by varieties of hybrid poplar (Thompson and Schnoor, 1996; Thompson et
al., 1998; and Schnoor, 1997).
48
Extensive studies on the phytoremediation of chlorinated solvents have been conducted at the
University of Washington (Newman et al., 1997: Newman et al., 1998; and Newman et al., 1999). In
recent years, researchers have begun to address the potential of phytoremediation to treat organic
contaminants other than TCE, including polynuclear aromatic hydrocarbons (PAHs) (Aprill and Sims,
1990; Pradhan et al., 1998; and Fiorenza et al., 2000) and polychlorinated biphenyls (PCBs) (Ferro et
al., 1994). Table 1.4 lists the organic contaminants that have been reported to be degraded more rapidly
in rhizosphere soil than in unplanted soil with phytoremediation.
2.4.1 Modeling Phytoremediation: Previous Work
Mathematical models of using plants in bioremediation/plume control are helpful for assessing the
practical implications of phytoremediation. Simulation models with some assumptions help to predict
the feasibility of proposed phytoremediation schemes. Knowledge of the groundwater hydrology, soil-
water fluxes, site geological characteristics, contaminant phyrotoxicity, and environmental factors are
critical in modeling plant-based bioremediation, (Trapp 2004).
Researchers have developed models to study movement of water in vegetated soils under tile
influence of evapotranspiration (Feddes et al., 1975, Neuman et al., 1975, Marino and Tracy, 1988).
Marino and Tracy (1988) proposed and verified a macroscopic root-soil water flow model that
simulated the movement of water through a vegetated environment. The model includes processes
such as water storage effects in the root and limiting and wilting root-water potentials that affect the
plant’s transpiration rate. Models developed to study the fate and transport of contaminants in the
presence of vegetation are relatively limited (Briggs et al., 1982, Boersma et al., 1990, Davis et al., 1993,
Trapp, 1995).
In one of the most used research articles in the area of plant uptake, based on studies with barley
plants, the uptake of several organics in homologous series, Briggs et al. 1982 proposed relationships
for RCF and TSCF based on linear regression with log Kow values. They found that compounds with
partition coefficient values of about 100 (log Kow = 2) show a maximum translocation into the
transpiration stream of the plant; typically this is as much as 80% of soil-water contaminant
concentration. They also cited examples of compounds that deviate from the predictive transpiration
stream concentration curve. Attempts to discern organic compound uptake and metabolism by plants
as compared to extent of rhizosphere biodegradation, are presently a challenge to plant physiologists,
environmental engineers, and microbiologists.
49
Briggs et al. (1982), defined two terms, root concentration factor (RCF) and transpiration stream
concentration factor (TSCF), to mathematically represent the adsorption and translocation of the
organics in plants. RCF is defined as the ratio of the contaminant concentration in the roots to that in
the soil-water. Whereas, TSCF is defined as the ratio of the contaminant concentration in the
transpiration stream to that in the soil-water.
Main point of value for Briggs et al. (1982) research paper was the declaration that TSCF is
independent of concentration of the external solution. Also they defined the RCF as the concentration
in roots over the external solution concentration, which seems to imply that it means the mass in the
roots and not necessarily sorbed – but later references to RCF both use the “in” term but go on to
clarify that the mass they are referring to is that sorbed to the outside of the roots (Burken and
Schnoor, 1997, 1998) or to the endodermis – an internal part of the cortex in the roots that separates
the xylem in the roots from everything outside (Trapp, 1995). Either way, this sorbed concentration is
stuck in the roots, and not subject to uptake by the plant with the transpiration water stream.
Boersma et al. (1990) modeled the passive and active uptake of xenobiotic chemicals by a
compartmental representation of the physical and chemical processes in terrestrial plants. They also
accounted for movement of water and organic nutrients within the plants. Models considering active
and passive processes for uptake of contaminants interacting with roots and shoots have also been
studied (Trapp and McFarlane, 1995).
Trapp et al. (1995), developed generic one-compartment model for uptake of organic chemicals by
foliar vegetation. It presents equations that indicate equilibrium between the two phases, and also
operate on the principle that using TSCF values based on hydroponic experiments for modeling plants
in soil. The model itself generates a linear differential equation of first order with concentration in the
plant leaves being the only variable with respect to time.
Partitioning is assumed to be equilibrium between soil and porewater, flux is coming in from soil
(using the TSCF equation from Briggs et al., 1982) and air, and degradation is occurring within the
plant. Among the simplifications the model has, steady environmental conditions is one of the most
significant. They also generated an equation showing time to reach steady state, with values in the
range of a week or two for two examples calculations.
The result includes the effects of uptake from air and degradation internally, and therefore isn’t
necessarily representative of a value for the uptake from soil process, which seems, based on other
50
researches, to be much quicker (Burken and Schnoor, 1997, 1998). They relied on TSCF data from a
Kow based formula from (Briggs et al. 1982) that used hydroponic data, but apply it to soil.
Behrendt et al. (1995) created a dynamic numerical model from the perspective of the soil and as a
dynamic uptake model that uses TSCF. They express the equations mainly in terms of the bulk soil
concentration, and give partition equations to soil water (which they assume is in local equilibrium), but
not air, although they don’t really say whether this is a saturated or unsaturated case. They also did an
analytical model, also in terms of total soil concentration, and derive some interesting equations for soil
concentration and total soil mass over time due to the effects of plant uptake, in-situ biodegradation,
and leaching. They concluded that the maximum of the pesticide root uptake as a function of sorption
parameters depends on the degradation rate of the chemicals in the Autumn scenario, but almost not in
the Spring scenario.
In 1995, Narayanan et al., conducted experiments and mathematical modeling to get at how alfalfa
plants affect biodegradation not only by uptake, but also how plants influence the hydrology and
geochemistry of the soil to increase the biodegradation that is going on in the soil. The experimental
model consists of a chamber of a two U-shaped channels packed with fine sandy soil collected from
near a landfill. Alfalfa plants were grown in the channel under laboratory conditions for nearly two
years. The water fed to the plants in one channel was contaminated with toluene, and the other channel
with phenol solution at different concentrations. The contaminant concentrations in the groundwater
were monitored at sampling wells located along each of the channels. In the mathematical model, they
used the variably saturated 1-D model used by Davis et al., 1998. The root-soil water flow model and
the variably saturated contaminant degradation models were solved simultaneously using a Galerkin
finite element method.
Burken, and Schnoor (1997) investigated the uptake and metabolism of atrazine by poplar trees. In
their research, they applied the TSCF concept to a dynamic uptake model, which is based on Trapp et
al. (1995). The dynamic mathematical model was intended to simulate the experiments only and was
not intended to be a general model. Their use of the essentially steady-state concept of TSCF in a
dynamic uptake model implies that doing so is reasonably accurate – i.e. that plants return to steady-
state quickly enough after a perturbation that use of such an approach isn’t particularly inaccurate. This
paper also presents the equations to relate the atrazine concentration in porewater and both RCF, and
TSCF:
51
[ ] [ ] [ ] [ ]
[ ] [ ] [ ]⎟⎠⎞
⎜⎝⎛ −−
−−−−−=⎟⎠⎞
⎜⎝⎛
RCFAtraAtrak
VAtraTTSCF
AtrakAtrakAtrakdt
Atrad
RSWS
W
WAtra
WWWWWWW
321
………. (2.20)
[ ] [ ] [ ] [ ] [ ]LLLLLLL
RL AtrakAtrakAtrakVAtraT
dtAtrad
321 −−−−=⎟⎠⎞
⎜⎝⎛ ………. (2.21)
They apply the TSCF concept to porewater in soil, and turn around and plug in TSCF data for
atrazine derived based on the Kow and equation for TSCF derived from hydroponic experiments (Briggs
et al. 1982), so it appears that hydroponic data can be used for soil scenarios, although they don’t
explicitly say they are doing so, nor give reasons why doing so is valid. (Trapp et al, 1995) also does the
same thing.
This paper indicates significant atrazine degradation in the roots, which means that TSCF data
generated by measuring the flux from the plant plus remaining mass in the plant would miss a
potentially significant amount of mass taken up by the plant and metabolized and therefore
underestimate the TSCF and therefore underestimate uptake. This would vary by compound and
means that where in-plant degradation is significant, TSCF data may not be accurate. The model
developed in this paper accounts for metabolism within the plant as a separate term, after the mass has
been brought into the plant transpiration stream via the TSCF factor.
Burken and Schnoor (1998) tried to predict the relationships for uptake of organic contaminants by
hybrid poplar trees. In their article better characterized the TSCF conceptual model, which appears to
be an empirical model of steady-state flux. The paper presents the TSCF and RCF
data/curves/equations from (Briggs et al. 1982), and adds their own experimentally derived data, and
creates new curves/equations for comparison.
The paper indicated that steady-state will be quickly achieved and TSCF accounts for mass
transpired and mass in plant without referring to the mass transformed in plant which would vary by
contaminant. It also provides equation (2-21):
solution bulk
stream iontranspiratC
CTSCF =
solution bulkCTransTSCFuptake ××= ………. (2.22)
52
( ) ( )2,1,2121 21
t solution bulkt solution bulktttt CCTransTSCFuptake +××= −−
The research indicated that these results are from hydroponic experiments in the absence of soil
sorption processes.
Aitchison et al., 2000, did both hydroponic and soil experiments. Hydroponic results are presented
in the familiar TSCF format, and the following results were obtained:
• 30-79% (average = 54%) of the dioxane mass had been removed from the planted reactors • 10% removed from the excised tree reactors • 8% removed from the unplanted control • Concentration of 1,4-dioxane remained relatively constant in all reactors, indicating that the
compound may be freely diffusing into the plant via water osmosis.
The results indicate that degradation of 1,4-dioxane by indigenous root-zone microorganisms is
minimal in comparison to plant uptake. The majority of 1,4-dioxane taken up into the plant was
volatilized (average = 77%), with the remaining mass concentrated primarily in the stem. Rapid uptake
of 1,4-dioxane by hybrid poplar trees makes phytoremediation appear as an attractive alternative at
dioxane-contaminated sites. Further research will examine poplar removal of 1.4-dioxane from
contaminated soil, (Aitchison et al., 2000). Although the dioxane results do not fit the fitted equations
from other authors, reason is suggested that there are ways that hydrophilic substances can be taken
into roots without having to be directly transported across via lipophilic mechanisms at the bilayer. Soil
results are not presented in TSCF form, even though soil moisture was kept consistently at nearly field
capacity during the experiments.
In 2001, Landmeyer presented direct and indirect methods to monitor groundwater use by Poplar
trees:
1- Measuring groundwater level changes using monitoring wells, which recorded a maximum
decline, by using sensitive pressure transducers that can resolve up to 0.01 ft or greater change
in water level.
2- Monitoring the water pressure: The reduction in groundwater levels near the surface of the
water table can lower water pressures beneath the trees throughout the entire saturated
thickness of the aquifer, hence, a vertical flow component can exist in saturated zones at
depths greater than root penetration.
53
3- Measuring the downgradient groundwater flux. The reduction of groundwater flux will
indicate the poplar trees usage of water.
4- Measuring the contaminant mass flux (Q×C) upstream and downstream the poplar trees. The
difference will estimate the plant contaminant uptake.
Chiou et al., 2001 presented a passive transport model for roots uptake. For a contaminant at a
location within the plant, local equilibrium is assumed to exist between plant water phase, and various
plant organic components. This article defines uptake as mass that is taken up by the plant from soil
and stays in the plant (i.e. doesn’t volatilize out the leaves, etc). The focus is what mass will remain
when, for instance, the plant is eaten.
The more water-soluble compounds (Kow ≤ 100) would quickly equilibrate their concentrations in
plant water with those in pore water. Some of this plant water is mobile as part of the transpiration
stream, and volatile compounds would therefore attain a steady-state flux out the leaves. A plot of
total mass stored in the plant would level off fairly quickly for both volatile and non-volatile
compounds. A plot of total mass removed from groundwater would increase linearly for volatile
compounds, similar to the linear plots shown in the TSCF experiment papers. A plot of total mass
removed from groundwater would be identical the plot of mass stored in the plant for non-volatile
compounds, and would therefore level off quickly. This would make the TSCF model not apply to
non-volatile compounds. All of this ignores plant growth, which would tend to make the plot of total
mass in the plant (and therefore also the plot of total removal from groundwater for non-volatile
compounds) increase slightly instead of leveling off completely, and make the plot of total removal of
groundwater for volatile compounds curve slightly upward instead of increasingly linearly.
The more fat-soluble compounds (Kow ≥ 100) would behave the same way as the water soluble ones,
except that attainment of equilibration concentrations in plant lipid with those in pore water would
occur more slowly. This is because the concentration of the fat-soluble compounds in the
transpiration water flux would be less (due to their lipiphilicity), and their sorptive tendencies will be
greater (again, due to their lipiphilicity). Unlike with the water-soluble compounds, the flux of fat-
soluble compounds out the leaves would remain minimal for a noticeable amount of time, even for
volatile compounds, during which the plant lipid would be getting saturated with that particular
compound. For both water and fat-soluble compounds, the approach to equilibrium occurs faster with
faster transpirational water flux.
54
The value for TSCF should account for the flux of both types of compounds, and therefore vary
between different plant species (based on, among other things, varying lipid content) and different
compounds (based on, among other things, varying lipiphilicity). Although a volatile fat soluble
compound could take a significant amount of time to reach equilibrium and start to appear in the water
transpired out the leaves, the uptake via the roots should stay constant with time, and if TSCF values
are calculated by adding together mass volatilized out the roots to mass contained within the plant, they
should be accurate whether or not the compound has come to equilibrium with the plant lipid phase.
That is, unless a significant fraction of the compound metabolizes in the plant, (such as for barley,
compounds with log Kow < 3), equilibrium is largely reached within 2 days, whereas for compounds with
log Kow > 3 it is a bit more complicated.
The model points out that uptake of organics is related to organic matter content of the soil –
meaning if you dump a bunch of organics in the soil, more will be taken up by plants if they’re growing
in a mineral soil than if they’re growing in muck. However, the reason this is true is that more will
partition to the water in the mineral soil, and because we’re going to be specifying water concentration
as the driver getting mass into the roots, and we’ll be accounting for sorption (which in theory should
account for organic matter partitioning via a Koc×foc type of equation). In summary, they stated that for a
contaminant at a location within the plant, local equilibrium is assumed to exist between plant water
phase and various plant organic components; however, the local concentration may or may not be in
equilibrium with external water, (Chiou et al., 2001).
Kijune et al., 2001 conducted a study to investigate the plant contamination by organic pollutants in
phytoremediation. This study modeled the interaction by considering two-compartments model: root
compartment in interaction with soil and groundwater, and shoots compartment in interaction with air
and the root compartment and models the dynamic uptake of two organic, non-volatile compounds
that do metabolize somewhat in the plant. The uptake into the transpiration stream within both the
roots and shoots uses the familiar TSCF. The soil sorption to roots uses the familiar RCF equation of
(Briggs et al., 1982) after adding a kinetic part to it.
Kijune et al., 2001 presented plots indicating that the two compounds are sucked up into the plant
until they reach equilibrium with the soil water concentration in both roots and shoots, and then slowly
decline as degradation reduces the concentrations both in the plant and indirectly via equilibrium
adjustment in the soil. The results also showed that more lipophilic compounds reach equilibrium
more slowly than hydrophilic ones, and that at least with roots, they reach a higher concentration in
55
plant tissue. They don’t reach a higher concentration in shoots because of the filtering effect of the
TSCF. They also considered degradation by microbes and sorption in the rhizosphere in the model,
and estimated transpiration water flux using a water stress index. The model also indicated that: 1)
uptake from soil air is negligible even in vadose zone, 2) non-volatile compounds do not leave the
shoots other than by degradation/metabolism, 3) flux downward via phloem is dwarfed by flux
upward via xylems.
In 2002, Ying presented a phytoremediation model for plant uptake and contaminant transport in
the soil-plant-atmosphere continuum. They took the CTSPAC model that was developed to model
coupled transport of water, heat, and solutes in the soil-plant-atmosphere continuum in 1-D and
adapted it to phytoremediation.
The model control volume is the soil vadose zone in this case, and they included both advection
and diffusion into roots from what appears to be soil pore water in their equations, and they present
the equations, although there may not be enough info from this article alone to apply them confidently.
Either way, their model seems more rigorous and complicated than the TSCF model
Freyer and Chistopher, (2003), compared a series of equilibrium (regression), steady state, and
dynamic models for uptake of organic compounds by herbaceous plants from soil (as well as air). They
attempted to validate the models with independent data from both hydroponic and soil growing
conditions. They selected three dynamic, three regression based, and three steady-state models making
a total of nine models for comparison. The results indicated that dynamic models offer performance
advantages for acute exposure durations and for rapidly changing environmental media.
Equilibrium/steady-state regression-based models perform better for chronic exposure durations,
where stable conditions are more likely to exist.
Thoma and Wolf, 2003 presented a Mathematical Model of Phytoremediation in which, detailed
site-specific information is not needed for Petroleum-Contaminated Soil. The model took into account
the root length density but not uptake by plants. The model was equilibrium mass-balance four
compartments representing the root itself, the rhizosphere, a decaying root zone, and a non-root-
influenced zone (the bulk soil). The model takes into consideration the root growth and thus the
corresponding volumetric changes in the other compartments and describe the rate of growth and
decay of the root biomass as a function of time.
56
57
Many authors have investigated models for solute transport simulation in groundwater. The most
popular models available are:
1- MOC is the USGS 2-D Solute Transport and Dispersion in Ground Water by L.F. Konikow
and J.D. Bredehoeft.
2- MT3D - Modular Three-Dimensional Transport Model,: MT3D is capable of modeling
advection in complex steady-state and transient flow fields, anisotropic dispersion, first-order
decay and production reactions, and linear and nonlinear sorption.
3- 3DFEMFAT - 3-D Finite-Element Model of Flow and Transport through Saturated-
Unsaturated Media.
4- ANALGWST (DG) - Version: 1.1 last updated 1996/04/03: A set of programs that calculate
analytical solutions for one-, two-, and three-dimensional solute transport in ground-water
systems with uniform flow.
5- BIOMOC (DOS/DG/SGI/Sun) - Version: 1.0 last updated 1999/03/10: A multispecies
solute-transport model with biodegradation.
6- HST3D (DOS/DG/Sun) - Version 2.2.11 last updated (Mar. 5, 2004): Three-dimensional
flow, heat, and solute transport model.
7- SUTRA and related programs: 2D, 3D, variable-density, variably-saturated flow, solute or
energy transport, and others.
8- SEAM3D, “Sequential Electron Acceptor Model 3-Dimensions”: a numerical model for
subsurface solute transport with aerobic and sequential anaerobic biodegradation. SEAM3D is
based on the numerical model MT3DMS. Extending the simulation beyond just the solute
movement, SEAM3D is also taking into account both chemical reaction and electron acceptor,
and sequential aerobic/anaerobic biodegradation.
The above models, including SEAM3D are not having a package for solute plant uptake, and root
sorption, which would be useful for determining the potential of using phytoremediation at
contaminated groundwater sites. Table 2.6 lists a comparison of most popular plant uptake models.
Table 2.6. Contaminant fate transport models comparison. Reference Control Volume Phases Type Abstract
Briggs et al., 1982 Soil/Plant (Barley) Soil Water/Plant Equilibrium Experimental/empirical Measured equilibrium concentration in soil and plants.
Burken J. G. and J. L. Schnoor. 1998
Soil/Plant (poplar trees) Soil Water/Plant Equilibrium
Experimental/empirical Fate of 12 organic compounds in poplar trees.
Marino and Tracy, 1988 Root hair Soil water/roots Mass balance
Variably-saturated flow model via a root extraction term that is a function of the water pressure gradient across the root-soil interface as well as soil and root parameters.
Trapp, and Matthies, 1995
Roots/shoots (2 compartments)
Soil water, plant, air Mass balance
Applicable to grass and green fodder. Processes considered are: translocation to shoots, gaseous depositions on leaves, volatilization from leaves, metabolism and degradation processes, dilution by exponential growth.
Trapp et al., 1995 Whole plant parts (root, stem, and fruits)
Soil water, plant, air Mass balance
Assumptions: 1- There are no transport processes except the passive processes of diffusion and advection. 2- The partition between plant tissue and aqueous solution is driven by the lipid and water content of the plant and the lipophilicity of the chemical (expressed as Kow).
Behrendt et al., 1995 Soil/Plant Soil water, plant Equilibrium, 1-D
Homogenous partially saturated soil, constant 1-D vertical leaching (no diffusion/dispersion), time constant and depth constant root water uptake rate, equilibrium distribution of chemicals between soil matrix and soil water.
Narayanan et al., 1995 Soil/Plant Soil water, plant Experimental/Equilibrium Used Alfalfa plants for the two-years experiment.
Chiou et al., 2001 Soil/Roots Soil water, roots Equilibrium Passive transport model for roots uptake. For a contaminant at a location within the plant, local equilibrium is assumed to exist between plant water phase, and various plant organic components.
Trapp, 2000 Soil, Roots, plant cell Soil water/plant Equilibrium and dynamic
steady-state. ES
The model approach combines the processes of lipophilic sorption, electrochemical interactions, ion trap, advection in xylem and dilution by growth.
58
59
Table 2.6. Contaminant fate transport models comparison, continued. Reference Control Volume Phases Type Abstract
SWMS_3D Partially saturated soil Soil/water Mass balance –Dynamic 3-D
(no plant uptake)
SWMS-3D a computer program for simulating 3D water flow and
solute transport in variably saturated media. It can be integrated with
UNSAT-H to simulate plant water uptake (but with no partition factor,
i.e. it assumes 100% of chemical mass is transferred to the plant, or,
TSCF=1.0).
Trapp et al., 1998, CemoS (Chemical exposure model System)
Soil water, plant, ait
Soil water, plant, air Equilibrium/mass balance
CemoS was developed for the exposure prediction of hazardous
chemicals released to the environment. Nine different models were
implemented involving chemicals fate simulation in air, water, soil and
plants after continuous or single emissions from point and diffuse
sources. Scenario studies are supported by a substance and an
environmental database.
Paterson and Mackay, 1994
Soil, plant (root, stem, and foliage), and air
Soil/water, plant, and air
Mechanistic/mass balance/dynamic ES
A three-compartment model of chemical transport and transformation in a plant exposed to soil and air.
SWAP (Soil, Water, Atmosphere and Plant)
Soil water, roots in partially saturated soil, and air.
Soil /plant/air Mass balance –Transient Dynamic 1-D
Based on Richards’ equation, SWAP simulates vertical transport of water, solutes and heat in unsaturated/saturated soils. The program is designed to simulate the transport processes at field scale level and during entire growing seasons.
* ES = Environmental Segment.
60
2.5 Phytoremediation Technical Considerations
Several criteria should be considered before phytoremediation of organic contaminants in the
rhizosphere is selected as an appropriate treatment option for a particular contaminated site. These
criteria are related to the chemical and environmental characteristics important to microbial
degradation in general as well as the characteristics (limitations) of the vegetation specifically. The
mechanism of vegetation uptake of organic pollutants is governed by the chemical and physical
properties of the pollutant, environmental conditions, and the plant species, (ITRC, 1999).
Vapor pressure reflects the volatilization potential when the chemical is not yet dissolved in a
groundwater system. Water solubility is an indication of the extent to which the compound can
dissolve into the water phase. The Henry's Law constant is an indicator of the equilibrium distribution
of a compound between water and air. The organic carbon water partition coefficient (Koc) is a
reflection of the compound’s tendency to sorb to the organic carbon matrix within soil systems. The
organic carbon sorption will retard the migration of the compound. The octanol/water partition
coefficient (Kow) of an organic contaminant is an important parameter to assess when considering the
potential of phytoremediation for cleanup (Burken and Scanner, 1997).
The Kow is related to observed root uptake and translocation of organics within plants. Hydrophobic
compounds such as PAHs (log Kow greater than 3) are not translocated to above ground plant tissues
(shoots and leaves), (Aprill and Sims, 1990). Uptake from soil through plant roots is the predominant
pathway of accumulation for organic compounds that have high water solubility, low Henry’s Law
constants, and low Kow values.
Hydrophobic chemicals log Kow > 3.5) are expected to be sorbed strongly to soils and not
bioavailable to plants for translocation. Moderately hydrophobic chemicals log Kow = 1~3.5) are
expected to be taken up by plants and metabolized, volatilized, or incorporated into plant tissues as
non-extractable bound residue. Hydrophilic chemicals log Kow < l) are not expected to be taken up or
sorbed by plants (Schnoor, 1997). The phytoremediation decision tree is presented in Figure 2.14,
ITRC 1999.
61
Decision Tree for PhytoremediationGroundwater
YES Will the climate support the proposed plants NO
Is time or space a constraints
Is the contaminant physically within the range of the proposed planttypically less than 10-20 feet bgs for Salix species - willows, cottonwoods, poplars)?
Will the water be mechanically pumped and applied to the Phytoremediation system?
Will the state regulations allowthis type of phytoremediation?
Will the plants be used for hydraulic control only (prevent groundwater from
reaching the contaminated zone)?
Is the contaminant at phytotoxic concentrations(this may require a greenhouse dose-response test)?
Will the rhizosphere microbes and plant-exuded enzymes degrade the targetcontaminants in the rhizosphere and will the metabolic products be acceptable?
Is the log Kow of the contaminants or metabolicproducts between 1 and 3.5 (will uptake occur)?
Will the plant degrade the contaminant after uptake and are
the metabolic products acceptable?
Will the plant accumulate the contaminantor metabolic products after uptake?
Will the plant transpire the contaminant or metabolic products? Is the level of accumulation acceptable
for this site throughout the growth of the plant?
Is the quantity and rate of transpirationacceptable for this site?
Can engineering controls make it acceptable?
Can controls be put in place to prevent the transfer of the contaminant or metabolic
products from a plant to human/animals?
Can the contaminant or metabolic productbe immobilized to acceptable levels?
Is the final disposition of the contaminant or metabolic products acceptable?
Does the plant material constitute a waste if harvested?
Can the plant waste be economically disposed?
Phytoremediation has the potentialto be effective at the site
Phytoremediation is not an optionat the site; consider other options
YESNO
YESNO
NO
YES
NOYES
YESNO
NOYES
YESNO
YESNO
NO
YES
YESNO
YESNO
NO
YES
YESNO
YESNO
NOYES
NOYES
NOYES
NO
YES
YESNO
Figure 2.14. Decision tree for phytoremediation.
62
2.5.1 Advantages of Phytoremediation
Phytoremediation is cost-effective. As a stand-alone solution, phytoremediation costs between one-
tenth and one-third that of conventional remediation technologies. Both capital costs and operating
costs of phytoremediation are minimal. As an adjunct to conventional remediation methods,
Phytoremediation reduces both cleanup time and operations and maintenance costs. The cost of
phytoremediation is 10-50% of the cost of mechanical, thermal, or chemical treatments (Flathman and
Lanza, 1998). Phytoremediation is a permanent in situ solution. Most conventional methods result in
the transfer of contaminants from one medium to another or from the site to a landfill, merely
postponing a permanent solution.
2.5.2 Limitations of Phytoremediation
Phytoremediation technology application is limited by a number of factors despite its diversity. The
limitations of phytoremediation are that contamination must be shallow, the site must be a large
enough to apply agronomic techniques, there must be sufficient remedial time, and its effectiveness is
affected by contaminant variability, weather variability, animal and insect damage, and the presence of
toxic chemicals and salt. Phytoremediation can only work at sites that are well suited for plant growth.
This means that the concentration of pollutants cannot be toxic to the plants, and the pollution cannot
be so deep in the soils or groundwater that plant roots cannot reach it. As a result, phytoremediation
may be a good strategy for sites conducive to plant growth with shallow contamination, it may be a
good secondary or tertiary phase in a treatment train for highly polluted sites, or it may not be a viable
option for a site. A brief comparison between advantages and limitations of phytoremediation as a
remedial option is listed in Table 2.7.
2.5.3 Costs of Phytoremediation
In the United States the costs of remediation is astronomical, with an estimate of surpassing 700
billion dollars for the tens of thousands of contaminated sites that need to be cleaned-up (Revkin,
2001). So far, 410 Superfund Sites (32%) on the National Priority List (NPL) have been remediated of
hazardous waste to levels safe for human health and the environment. The most common technologies
used in these clean-up projects was excavating and removing hazardous soil and solid waste (45%),
covering the landfill with a protective cap (39%) and pumping and treating contaminated groundwater
(34%). These technologies are very costly. Cost estimates for excavation and disposal range from
63
$270.00 to $460.00 per ton depending on the nature of hazardous materials and methods of excavation
Approximate industry costs for capping a contaminated site are $175,000 to $225,000 per acre
(www.frtr.gov). Not only are these two technologies costly they do not eliminate the contamination,
but move the waste in an area that has no access to the public. Actual costs of pumping arts treating a
chlorinated solvent. volatiles, and selenium contaminated site was $27,600,000, which corresponds to
$23.00 per 1000 gallons of groundwater extracted and $64.00 per pound of contaminant removed.
Table 2.7. Major Advantages and Disadvantages of the Phytoremediation Process.
Advantages Limitations
Less soil disturbance compared to conventional methods
Restricted to sites with shallow contamination within the roaring zones of remediating plants.
Reduces by up to 95 percent the amount of waste to be landfill
Slow reaction rates: Phytoremediation may take several growing seasons to clean up a site effectively, (Rock 1997).
Useful as an in situ and ex situ application Restricted to sites with tow contaminant concentrations Reduces the cost of remediation as compared with the cost of standard engineering methods
Harvested plant material from phytoextraction may to classified as a hazardous waste
Reduction in soil erosion, (Ecological Engineering 1998).
Remediating plant materials restricted by climatic and cite conditions
Cost-effective technology: can reduce the cost of clean up of a site to between one-third and one-hundredth of the cost of some existing remediation technologies (Boyajian and Devedjian 1997).
Seasonal constraints: Many climatic factors may influence the effectiveness of a phytoremediation system, such as rainfall patterns, wind duration at various seasons, etc. (ITRC, 2000).
Applicable to treat a wide variety of contaminants: Amenable to a variety of organic and inorganic compounds
Shallow, low/moderate levels of contaminant concentration: If the contaminant concentration is too high, the contaminant will be toxic to the plant species (Schnoor 1997).
Aesthetically pleasing: Plants and vegetation can clean up sites with minimal disruption to the local community (Boyajian and Devedjim 1997).
Large surface area required: Phytoremediation systems can require large surface areas of land, in order to completely remediate the contaminant (ITRC, 2000).
Permanent treatment solution: Phytoremediation permanently decreases the availability, toxicity, and concentrations of contaminants (Banks et al.2000).
Unfamiliar to regulators: up to 2000, regulatory standards for phytoremediation have not been developed. Therefore, regulators evaluate and approve the proposed phytoremediation applications on a site-by- site basis (Rock 1997).
In situ application avoids excavation: less secondary wastes are generated since the soils are not removed (Chappell 1997).
High public acceptance
Estimates of costs for phytoremediation of a one acre site, including site preparation, planting, and
removal (harvest) of plant material, range from $2000.00 to $5000.00 (Phytokinetics). US AEC
estimated that the cost for phytoremediation of one acre of lead-contaminated soil to a depth of 50-cm
was $60,000 to $100,000, whereas excavating and land filling the same soil was $400,000 to $1,700,000.
Growing a green crop on an acre of land can be completed significantly less (2-4 orders of magnitude)
than excavation and reburial (Cunningham, 1996).
64
One out of many success stories for phytoremediation will be presented next. A phytoremediation
company used sunflowers and Indian mustard to remediate lead-contaminated soil in Detroit. The lead
contamination was reduced by 43% with a project cost of $900,000. It was estimated that the costs of
hauling off the 5,700 cubic yards of lead-contaminated soil would have been more than a million
dollars (Revlon, 2001). Table 2.1 compares some of the costs of other remedial technologies to
phytoremediation.
2.6 Research Deficiencies
Most of the research studies in phytoremediation have focused on plant physiology (testing and
developing new plant species suitable for use in phytoremediation) and the effect of plants on
groundwater and soil pollution. Research on the uptake of water and solutes in the field is very rare.
The effect of plants on groundwater levels and the amount of water that can be extracted by plants by
evapotranspiration is on great interest. However, only a few attempts to predict water table drawdown
or to estimate the extent of aquifer zone potentially affected by ET from phytoremediation plantation
are documented in the literature, (Hong et al., 2001, and Mathews, et al., 2002). Solving the problem of
phytoremediation system effect on groundwater is critical for the design point of view, as for an
efficient capture of the groundwater plume, or for groundwater control purposes, the impact of PPS
should be known.
2.7. Research Aims
The need to model the plant uptake is important to monitor the remediation of contaminated soil
or groundwater. Focusing on the concept of capturing the groundwater in the contaminated area is not
enough to indicate that the plume is controlled. As been mentioned in the literature, the tendency of
contamination to be uptaken depends on Kow, and thus depends on RCF and TSCF.
From this point of view comes the objective of this thesis to develop a new transport package for
SEAM3D called the Phytoremediation Uptake Package (PUP) to incorporate contaminant mass loss
from groundwater due to sorption and uptake by plants. This new package accounts for both uptake
and sorption to plant roots, as well as uptake into the transpiration stream of plants. The new package
is designed to cross-over the Evapotranspiration package of MODFLOW which does not take into
consideration the TSCF and assumes that 100% of mass is removed with the transpired groundwater.
65
The new phytoremediation package will use the same input files used by MODFLOW except for
the ET files. SEAM3D/PUP will use its own input file for plant uptake which is similar to ET file but
involves the employment of TSCF. The PUP will use the same source/sink input files used by
SEAM3D, and it will have its own input file for root sorption. Plant uptake and root sorption
information will be saved in the file with extension (*.pup).
66
C h a p t e r 3
Model Development
3.1 Conceptual Model
In the subsurface, dissolved organic chemicals are known to be removed by the influence of the
root systems of phreatophytic plants by any one of three mechanisms:
1. Direct Transpiration (Uptake)
2. Root Sorption
3. Biodegradation
The affinity of a solute to be transpired into the root system of a plant is quantitatively represented
by the Transpiration Stream Concentration Factor (TSCF). The TSCF of any compound x is defined
as the ratio of the concentration of x in the transpiration stream to the concentration of x in the
saturated zone. The value of TSCF varies from 0 to 1.0 and depends upon the chemical properties of
the compound, (Schnoor 2002).
The Root Concentration Factor (RCF) is a parameter that is similar to the distribution coefficient
used in modeling sorption to aquifer solids. The RCF of any compound x is defined as the ratio of the
concentration of x sorbed to the root system to the concentration of x in the saturated zone, (Schnoor
2002).
Values of zero for TSCF and RCF indicate that a solute will not be transpired by or sorbed to plant
roots, respectively. Relatively large values of TSCF and RCF for a compound reflect a high affinity for
transpiration and sorption, respectively.
67
3.2 Mathematical Model
Model variables and governing equations for SEAM3D are not presented in this report. The
complete system of governing equations used in SEAM3D consists of coupled partial and ordinary
differential equations describing solute transport, biodegradation, biogeneration, microbial growth and
decay, and sorption, (Widdowson 2002). Boundary and initial conditions are user-specified and are
required to develop a complete mathematical model. The description of the mathematical model for
SEAM3D-PUP is limited to sink terms for direct uptake and root sorption.
3.2.1 Direct Uptake
As a starting point, consider the equation of mass balance for the concentration (Sls) of a volatile
organic compound (VOC) in the mobile aqueous phase:
( )t
CCqt
CRxCD
xCv
xi
lssi
biksourcej
iij
iii
i ∂∂θρ
∂∂θ
∂∂θ
∂∂
=+∂
∂−+⎟
⎟⎠
⎞⎜⎜⎝
⎛+− *
,sin/ ……. (3.1)
where θ = aquifer porosity [Lo]; xi = distance [L]; t = time [T]; Ci = aqueous phase concentration
[Mls L-3] for VOC i; vi = average ground-water velocity [L T-1]; Dij = tensor for the hydrodynamic
dispersion coefficient [L2 T-1]; = mass source-sink term for reactions and mass transfer [Mls
L-3 T-1]; ρb = bulk density of the aquifer [Ms L-3];
iksourceR ,sin/
iC = solid phase concentration [Mls Ms-1] for VOC i;
qs = volumetric flow rate per unit volume of aquifer (total) representing fluid sources (positive) and
sinks (negative) [T-1]; and = VOC concentration associated with the point source or sink [Mls L-3]. *iC
The term represents the combined rate of mass removal due to all fluid sources and sinks. In
the case of a point sink, the concentration is generally not specified, and the codes (SEAM3D and
MT3DMS) use . Evapotranspiration is considered an areal sink in which the rate of mass
removal is calculated using either a user-specified sink concentration (which is independent of the cell
concentration) or the codes use .
*isCq
ii CC =*
ii CC =*
68
The modifications to SEAM3D are based on the concept of the transpiration stream concentration
factor (TSCF) so that = concentration of the transpiration stream = TiC iiSτ . The TSCF parameter
will be an input parameter that can vary over space and is compound specific. The volumetric rate of
direct transpiration of groundwater from the saturated zone (QET) is calculated using the
Evapotranspiration Package of MODFLOW. In SEAM3D-PUP the term general groundwater
source/sink qs is replaced by the areally distributed fluid sink term (qET) calculated in SEAM3D, where
qET is calculate at each cell as QET divided by the saturated cell volume. Equation (1) is then written as
( )t
CCqCqt
CRxCD
xCv
xi
iiETisi
biksourcej
iij
iii
i ∂∂θτρ
∂∂θ
∂∂θ
∂∂
=−+∂
∂−+⎟
⎟⎠
⎞⎜⎜⎝
⎛+− *
,sin/ ……. (3.2)
As shown in Figure 2.4, the magnitude of QET in any model cell varies from 0 to a user-specified
maximum ET rate (QETM) and is dependent on the hydraulic head, calculated cell-by-cell as
( )( )[ ] ( ) ssETM
sETMET
sET
sETMET
hhdh for .......... fQd
dhhQQ
dhh for ......... Qhh for ......... QQ
≤≤−×=−−
=
−>=>=
0 ………. (3.3)
Where h = elevation of the water table calculated by the model [L]; hs = land surface elevation [L]; d
= root extinction depth [L]; and f = volumetric fraction of the roots in the saturated zone.
The rate of solute mass removal for any compound due to direct plant uptake per model cell
volume is expressed in terms of the TSCF and the solute concentration, volumetric rate of direct
transpiration, and total volume of the cell
( ) ETicell
ETTiuptake
ik qCTSCFV
QCR ==,sin ………. (3.4)
where Vcell = saturated cell volume.
69
3.2.2 Root Sorption
Sorption of contaminants in the rhizosphere will be defined using the concept of the Root
Concentration Factor (RCF = ), defined as the ratio of , contaminant concentration sorbed to the
roots (mass per root mass), to the contaminant concentration in solution. Because the RCF is an
equilibrium model, this approach enables the rate of mass to be quantified in terms of the aqueous
contaminant concentration. For application to the root system of phreatophytes, the sink term for
mass removal to the roots is linked to the level of ground water relative to the root depth.
irRiC
The governing transport equation is revised to include an additional term for sorption to the root
system:
( )t
CCqt
Cft
CRxCD
xCv
xi
is
RiR
biS
biksourcej
iij
iii
i ∂∂θρρ
∂∂θ
∂∂θ
∂∂
=+∂
∂−
∂∂
−+⎟⎟⎠
⎞⎜⎜⎝
⎛+− *
,sin/ …. (3.5)
where = root density per total volume of aquifer [MR L-3]; and f represents the fraction of the
root system in contact with groundwater, as defined above in the MODFLOW ET Package.
Rbρ
By representing sorption using equilibrium models, the concentrations associated with the roots and
with the aquifer sediment are expressed in terms of aqueous concentration variable using and
, respectively. Equation (5) is then simplified to:
iiCr
iid CK ,
( ) ( )t
CfrKCqRxCD
xCv
xiR
blsidSbisiksource
j
iij
iii
i ∂∂θρρ
∂∂θ
∂∂θ
∂∂
++=++⎟⎟⎠
⎞⎜⎜⎝
⎛+− ,
*,sin/ ……. (3.6)
or
( )t
CRCqRxCD
xCv
xi
iisiksourcej
iij
iii
i ∂∂θ
∂∂θ
∂∂θ
∂∂
=++⎟⎟⎠
⎞⎜⎜⎝
⎛+− *
,sin/ ………. (3.7)
where the term Rls is the retardation factor modified for the sorption unto roots, given by
θ
ρθ
ρ frKR
Rbiid
Sb
i ++= ,1 ………. (3.8)
70
3.3 Model Implementation
The SEAM3D-PUP is designed to simulate the effect of the first two mechanisms. The
Biodegradation Package of SEAM3D is ideally suited to simulate the influence of the root systems on
microbially-mediated mass transformation and degradation.
Figure 3.1 shows the conceptual model of the SEAM3D-PUP. Plants simulated using PUP have
root systems that reached the saturated zone and possess the ability to transpire water from the
saturated zone (i.e., phreatophytes) Dissolved species are subject to either direct uptake into the plant
transpiration stream or sorption to the root surfaces, or both simultaneously.
For initial testing purposes, the Source-Sink Mixing Package (SSM) and the Reaction Package (RCT)
of SEAM3D were modified to simulate the effects of plant uptake and sorption to roots, respectively.
For final implementation in SEAM3D, a separate Plant Uptake Package was created to simulate both
effects simultaneously.
The SSM can be utilized in SEAM3D when using PUP to simulate any groundwater source or sink
with the exception of evapotranspiration. This eliminates any model errors created by “double-
counting” direct transpiration by both the PUP and the SSM. SEAM3D-PUP can only be executed
when the MODFLOW ET Package is active. When simulating root sorption without direct
transpiration, the maximum rate of evapotranspiration should be set to zero in the MODFLOW ET
Package and input for the SSM must be included. Simulation of rhizosphere bioremediation will be
implemented using the SEAM3D Biodegradation Package without any anticipated changes to the code.
The SEAM3D-PUP flowchart is illustrated in Figure 3.2. The code starts after MODFLOW is run
to find the hydraulic head values and thus the cell flow rates. The groundwater flux and the ET
flowrate are calculated after using the ET package as a sink term in the groundwater flow equation,
(McDonald, and Harbaugh 1988).
71
QET
M
ETQ
hsh
d
Max
imum
Evap
otra
nspi
ratio
n
Slop
e =
dETM
Q ____
__
0
shh
d
(h - d)s
Land surface elevation (SURF)
C
CR
CT = (TSCF) C
= (RCF) CVR
h-(h -d)s
Vt
Figure 3.1. Conceptual model for the two main mechanisms simulated using the SEAM3D
Plant Uptake Package.
72
Input Data
TSCF RCF
Figure 3.2. SEAM3D-PUP flowchart.
SEAM3D-PUP TSCF package is reading
QET solved by ET package in
MODFLOW, thus it’s only using the cells
assigned in the ET package as the plant
uptake cells. It is often referred to the
phytoremediation area as the ET area.
TSCF=T
Read the values of QET (solved by MODFLOW ET package)
Input arrays for TSCF values for each species in each layer for each stress period
For the GW transport equation, add the term
( ) ETicell
ETTiuptake
ik qCTSCFV
QCR ==,sin
to the sink term.
Read the corresponding values for SURF,
and EXDP for the cells with RCF > 0.
RCF=T
Input constant value for root density and array for RCF for each species in each layer for each stress period.
Calculate( )[ ]d
dhhf s −−=
Calculate total retardation factor:
θρ
θρ frK
RRbiid
Sb
i ++= ,1
END
73
C h a p t e r 4
Model Testing and Verification
4.1 Verification of the Plant Uptake Package
To test and verify the SEAM3D Plant Uptake Package, a number of simulations were performed to
demonstrate model capabilities and both major mechanisms. Simulations include closed system-
models where concentration versus time is simulated and dynamic transport models. The test
problems were selected to cover a wide range of real-world conditions. In most of the tests, the
SEAM3D-PUP package is tested against the SEAM3D/MT3D-SSM, and/or SEAM3D/MT3D-RCT.
4.2 Plant Uptake
Initially, test cases were selected to verify the plant uptake separate from the root sorption for
closed and open of flow systems. For each test case, the concentration break- through values and the
solute mass values are recorded and compared to SEAM3D-SSM (TSCF = 1.0 only) and for variable
TSCF values. The test cases for the plant uptake verification are:
1. Closed flow system, single MODFLOW stress period
2. Closed flow system, multiple MODFLOW stress period
3. Flow and transport model
4.2.1 Closed System Model – Single Stress Period
Time-dependent test cases were devised, in which advection, dispersion, and source/sink terms
were negligible, so concentration changes depended solely on plant uptake. In each case, the model
domain represented a 100 × 100 × 10 m unconfined aquifer, divided into 100 cells with dimensions of
10 × 10 × 10 m for each cell. In generating the groundwater flow field, the water table was horizontal
(h = 8.0 m), so all values of iv were zero in the transport simulations. The starting concentration of
74
the single solute model was uniform (10 mg/L), and rates of evapotranspiration and recharge (0.01
m/d) were identical at all nodes, forcing the concentration gradients to be zero.
Figure 4.1 depicts the model domain. In this 10-day single stress period simulation, the volumetric
rate of recharge (100 m3/d) is equal to the volumetric rate of evapotranspiration using MODFLOW.
The problem was simulated using the SSM Package in SEAM3D and SEAM3D-PUP with a value of
TSCF = 1.0. Simulation results in Table 4.1 and Figure 4.2 showing concentration versus time show
an identical match between the two codes. These results were also verified with using a spreadsheet
model. Dissolved mass removed through plant uptake versus time is included in the plot.
The problem was repeated to investigate the effect of the input parameter TSCF on concentration
and mass versus time. As the TSCF decreases, the mass of the solute taken by the plant decreases, and
thus increases the concentration of the solute in groundwater, compared with that of TSCF =1.0.
Tables 4.2 and Figure 4.3 show concentration versus time for five values of TSCF ranging from 0.0 to
1.0. Dissolved mass versus time is included in the table. A comparison of total mass removed through
plant uptake at time = 10 days shows that SEAM3D-PUP is working correctly for values of TSCF <
1.0.
4.2.2 Closed System Model – Multiple Stress Periods
The first problem was repeated for two MODFLOW stress periods over a 20-day simulation
timeframe to confirm that no coding errors were present in SEAM3D. The problem was simulated
using both the SSM Package in SEAM3D and SEAM3D-PUP with a value of TSCF = 1.0. At time =
10 days, evapotranspiration and recharge are turned off in the second MODFLOW stress period,
effectively eliminating direct uptake in the transport model. As a result, no change in the dissolved
phase concentration during the second stress period is expected. Simulation results for mass removed
and concentration versus time are presented in Table 4.3 and Figure 4.4. These results verify that mass
removal ceases with a change in the maximum evapotranspiration rate input parameter using the
MODFLOW ET Package.
This problem was repeated for four stress periods over 40 days. In stress periods 1 and 3, the
maximum rates of evapotranspiration and recharge rate are equal (0.01 m/d). In stress periods 2 and 4,
evapotranspiration and recharge are turned off in MODFLOW. The results are presented in Table 4.4
and Figure 4.5.
75
4.2.3 Flow and Transport with Direct Uptake
In this test case, advection, dispersion, and source/sink terms were included using a steady-state
groundwater flow model. The objective is to first match the concentration and mass results using
SEAM3D-SSM and to demonstrate the capability of SEAM3D-PUP to simulate mass removal by plant
uptake for different values of TSCF.
The model dimensions were 200 × 100 × 10 m for an unconfined aquifer (Figure 4.6). The model
domain was divided into 100 cells in the x-direction, and 50 cells in the y-direction, with cell
dimensions of 2 × 2 × 10 m. Groundwater inflow is generated by injection wells at the left model
boundary using a constant flow rate, Qin = 0.4 m3/d, in each cell. A constant head (h = 8.0 m) was set
at the downgradient right model boundary.
An area of ET is selected to be 20 m wide, and 100 m width to cover the whole width of the model
(Figure 4.6). The ET rate (0.01 m/d/cell) resulted in a total ET flowrate (QET) equal to 0.4 m3/d for
each row of cells. Because inflow = outflow, the water table was horizontal downstream of the ET
zone. One MODFLOW stress period equal to 3,650 days was used, but the SEAM3D simulation was
time dependent.
The problem was simulated using SEAM3D with the SSM Package and SEAM3D with PUP for
TSCF = 1.0. The problem was also simulated using SEAM3D-PUP for ranging from 0.0 to 1.0.
Simulation results in Table 4.5 and Figure 4.7 showing mass removal through plant uptake versus time,
and in Table 4.6 and Figure 4.8 showing concentration versus time, show an identical match between
the two codes (for the case when TSCF = 1.0 where 100 % of solute mass is extracted by ET). The
concentration breakthrough curves are presented for three observation points (i, j, k) = (24, 45, 1), (24,
50, 1), and (24, 56, 1) in Table 4.6 and in Figure 4.8 (top).
For the open flow-system model, as shown in the closed system model, when TSCF decreased, the
total mass of the solute removed by the plant decreased (Figure 4.7). Consequently, the concentration
of the solute in groundwater increased relative to simulation results for TSCF =1.0 (Table 4.7 and
Figure 4.8, bottom plot). A comparison of total mass removed through plant uptake at time = 3650
days shows that SEAM3D-PUP is also working correctly for values of TSCF < 1.0 (data not shown).
76
4.3 Root Sorption
Test cases were selected to verify root sorption separate from plant for different flow systems. For
each case, the concentration break through values and the solute mass values are recorded and
compared. To simulate the root sorption package, the plant uptake part of the PUP package is set off
(TSCF = F), and the root sorption package, RCF is (T). The test cases for the root sorption verification
are:
1. Flow and Transport with Root Sorption (f = 1.0)
2. Flow and Transport with Root Sorption (f < 1.0)
3. Flow and Transport with Root Sorption (f = 1.0) and Aquifer Sorption
4. Flow and Transport with Spatially-Variable Root Sorption (f = 1.0)
For the simulation of sorption to roots, SEAM3D-PUP is patterned after the Chemical Reaction
Package (RCT) in SEAM3D for sorption to aquifer solids. The RCT Package calculate the aqueous
concentration change due to sorption to aquifer solids using dSb
e
Kn
ρ1 . In the PUP package the same
term for the effect of root sorption is (RCFfn
Rb
e
ρ1 ) ). By setting in the two
separate models (SEAM3D-RCT and SEAM3D-PUP), a comparison of solutions was obtained. The
similarity between the SEAM3D-RCT package, and SEAM3D-PUP package can be achieved by setting
in RCT equal to in PUP, and Kd in RCT equal to RCF in PUP, and setting f = 1.0.
(RCFfK Rbd
Sb ρρ =
Sbρ R
bρ
Different values for root bulk density, , and RCF are assumed to maintain a constant value of
the retardation factor, R, represented in the equation:
Sbρ
de
Sb
e
Sb K
nCC
nR ρρ
+=∂∂
+= 11
Where Kd is the distribution coefficient [L3M-1].
4.3.1 Flow and Transport with Root Sorption (f = 1.0)
The model has the same settings as described in Section 4.1.3 (see Figure 4.6), with a difference that
ET rate is set to be zero (no plant uptake is simulated). Also, the plants roots are assumed to be active
over the entire model domain to produce a homogenous constant retardation factor.
77
The test is achieved by using SEAM3D- RCT with different values of Kd to control the retardation
factor, while in SEAM3D-PUP package, the plant uptake is not active, (TSCF is set to F), and the root
sorption is active, (RCF is set to T). Note that when implementing PUP, the retardation factor due to
soil sorption can be simulated, but to simulate the root sorption only, Kd is must be set to be zero in the
RCT Package (see Table 4.8). Three values of the retardation factor were simulated in each model (R
= 1.0, 1.5, and 2.0). Table 4.8 lists the values of Kd and RCF used to generate the three test values for
R. For each case, input parameters for solid and root densities were 31750000mgR
bSb == ρρ .
The test results (concentration breakthrough) and a comparison with SEAM3D-RCT at the
observation point (i, j, k) = (24, 50, 1) is presented in Table 4.9 and Figure 4.9 (bottom plot). The
results for solute mass removal at the same observation point for different values of RCF and
retardation factor are presented in Table 4.10 and Figure 4.10.
For the three values of retardation factor, the simulation results using SEAM3D-PUP showed an
excellent match with SEAM3D-RCT. The results in Figure 4.9 (top plot) demonstrate that the plant
roots can have retardation effect on the solute transport. As the values of the input parameters (f, ,
and RCF) increase, the retardation factor increases, decreasing the contaminant velocity, with the
increase of (f, , and RCF).
Rbρ
Rbρ
4.3.2 Flow and Transport with Root Sorption (f < 1.0)
The model settings are the same as in 4.2.1 but with a value of f < 1.0, the volumetric fraction of the
roots submerged in groundwater, which is a function of the hydraulic head in each model cell. Figure
4.11 (top) shows the hydraulic head distribution for the MODFLOW, yielding an average h value of 8.5
m. For hs = 10.0 m, and d = 4.0 m, the average f is 0.625.
The simulation results in this test case are difficult to compare directly with SEAM3D-RCT used in
GMS because SEAM3D-PUP package with f < 1.0 has a spatially-varying retardation factor with a
different value of f (and R) in each cell. To set a model in GMS using SEAM3D-RCT with a constant
retardation factor (R = 2.0) that can approximate the results of SEAM3D-PUP, the value of
was used in the GMS, yielding a homogenous retardation factor of 1.62. This 81093.8 −×=dK
78
compares well with an spatially-averaged retardation factor (1.625) in the SEAM3D-PUP model
domain.
Table 4.11 and Figure 4.11 (bottom plot) show concentration breakthrough curves for both models
at the observation point (i, j, k) = (24, 50, 1) for a constant and variable value of f. Unlike the case for f
= 1.0, the results for SEAM3D-RCT and SEAM3D-PUP do not show an exact match, but the
concentration values are relatively close for f < 1.0.
4.3.3 Flow and Transport with Root Sorption (f = 1.0) and Aquifer Sorption
This test problem is used to validate the SEAM3D-PUP package if both soil sorption (RCT
Package) and root sorption (PUP) are working simultaneously. The model setup for this test case is
exactly as that used in test case (4.2.1) but with f = 1.0, and the effect of root and aquifer media
sorption are equally combined.
The root sorption is combined with soil sorption by setting Kd in the RCT package and RCF in the
PUP Package so that R = 2.00 where 50% is due to roots and 50% is due to aquifer matrix. The model
parameters are shown in Table 4.12. To achieve this effect, Kd was set equal to 7.143×10-8 and RCF
was set to 7.143×10-8 to give a total R = 2.0 (1+0.5 from soil + 0.5 from roots).
To compare the SEAM3D-PUP results with the SEAM3D-RCT results, a model is set up as in test
case 4.2.1 with Kd = 14.3×10-8, which will give a retardation factor = 2.0. The results for this test case
are shown in Table 4.13 and Figure 4.12, which showed exact match between SEAM3D-PUP and
SEAM3D-RCT (concentration breakthrough curves) at the observation point (i, j, k) = (24, 50, 1).
4.3.4 Flow and Transport with Spatially-Variable Root Sorption (f = 1.0)
This test problem is designed to demonstrate SEAM3D-PUP will work properly if root sorption is
variable in designated model cells rather than uniform over the entire model domain. This problem is
designed to demonstrate that the root sorption cells can be controlled by input arrays of root bulk
density, or root concentration factor, RCF for the active root sorption areas, or both. Rbρ
The root sorption is inactive over the entire domain except for column # j=46 to j=55 (10 cells)
and along the whole model width (see Figure 4.6). Both Kd (in the RCT Package) and RCF (PUP) are
set to equal 7.143×10-8 m3/g. In the PUP model, values for = 1750000 g/m3 for the root sorption Rbρ
79
area and = 0.0 for the rest of the model area. The result is a model where R =2.0 in the root
sorption area and R = 1.5 for the rest of the model.
Rbρ
The results in Table 4.14 and Figure 4.13 show concentration breakthrough curve results for three
observation points using SEAM3D-PUP (spatially-variable R) and SEAM3D-RCT for a uniformly-
distributed retardation factor (R = 1.5). At the upgradient side of the phyto zone ((i, j) = (24, 45)) a
comparison of the breakthrough curves shows a nearly exact match. Downgradient, at the center and
end of the phyto zone, (i, j) = (24, 50) and (24, 56), respectively, the impact of the roots is apparent
with the increasingly delayed breakthrough relative to the SEAM3D-RCT results with uniform
sorption. These results demonstrate that the SEAM3D-PUP package is working properly if root
sorption is only active over a designated area of the model.
4.4 Direct Uptake and Root Sorption
For all the previous test cases, either plant uptake (TSCF) is active and root sorption (RCF) is
inactive or vice versa. In this test case, plant uptake is combined with root sorption. Two test cases are
presented:
4.4.1 Flow and Transport with Plant Uptake and Root Sorption (in the ET area
only)
The model setting is shown in Figure 4.6, which is the exact test case as 4.1.3, with one exception
Root sorption was added in the ET area only, from column # j=46 to j=55 (10 cells) and along the
whole model width (50 cells in y-direction). The values of Kd and RCF are both equal to 7.143×10-8
m3/g, which will give a retardation factor equals to 2.0 in the ET area, and 1.5 for the rest of the model
area. Two values of TSCF are selected: TSCF = 1.0 and TSCF = 0.5.
The results are presented in Table 4.15, Table 4.16 and Figure 4.14. The results show that low
values of TSCF is indicating low solute mass taken by the plants, and thus increased concentration. The
only case where the results of SEAM3D-PUP can be compared to SEAM3D results with the SSM and
RCT Packages is when RCF = 0.0 (no root sorption) and TSCF = 1.0. The comparison yields an exact
match.
80
4.4.2 Flow and Transport with Spatially Distributed RCF and Plant Uptake
The model setup for this problem is shown in Figure 4.15. This test case is different from case
4.2.1 in that the root bulk density and RCF are spatially variable in two different regions as shown in
Figure 4.15. The net retardation factor is 2.0 for the entire model domain where 50% of it comes from
soil, and 50% from roots. The retardation factor is calculated from the equation de
Sb K
nR ρ
+= 1 . The
plant uptake (QET) is active only in the middle region and equals to 0.01 m/d (from cell # 46 to cell #
56).
The results from SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (Tables 4.17,
Table 4.18 and Figure 4.16) show an excellent match. This verifies the capability of SEAM3D-PUP in
dealing with different combinations of plant uptake over a specified area with root sorption spatially
changing over the model area.
4.5 Conclusions
A new code for plant uptake and root sorption simulation in saturated unconfined aquifer is
presented. The new code named (SEAM3D-PUP) is verified by using hypothetical models with
different settings. The different model settings included closed system models with single and multiple
stress periods, flow and transport with direct uptake and root sorption spatially distributed over the
model domain. In each model run the results of the solute mass in the aquifer, and the mass removal
by ET (sinks term), and the solute dissolved concentration at specific observation points are compared.
In case of root sorption simulation, the retardation factor calculated from SEAM3D-RCT package, and
SEAM3D-PUP are compared. The new SEAM3D-PUP module demonstrated identical agreement
with SEAM3D-SSM and SEAM3D-RCT packages for plant uptake and root sorption simulations for a
wide range of model settings.
SEAM3D-PUP will enable modelers to simulate the effect of a phytoremediation system with
poplar trees on solute transport, in which TSCF and RCF can be incorporated. Previously, this has
been a limiting constraint because MT3DMS does not consider TSCF in the Source/Sink Mixing, SSM
package or RCF in the Reaction, RCT package. The only way to take TSCF into consideration when
modeling contaminant transport in MT3DMS is to assume it is equal to 1.0, and to simulate the effect
of the root sorption is to increase the sorption capacity of the aquifer layers in the areas of a
81
phytoremediation system. Another limitation of simulating TSCF and RCF in MT3DMS is that they
are not a function of the saturated thickness or the aquifer water table. TSCF and RCF are inherent
physical/chemical properties of the solute compound in groundwater which makes the previous
models inflexible it terms of dealing with different contaminant transport simulations.
82
Table 4.1. Comparison of concentration versus time from SEAM3D-PUP to both an exact Solution and SEAM3D-SSM for the closed-system, single stress period model.
Concentration Time, Days
SEAM3D-SSM Exact Solution SEAM3D-PUP
0.0 10.00 10.0000 10.0000 2.0 9.95 9.9500 9.9500 4.0 9.90 9.9003 9.9002 6.0 9.85 9.8507 9.8507 8.0 9.80 9.8015 9.8015 10.0 9.75 9.7525 9.7525
Table 4.2. Simulation results for mass removed by direct uptake and dissolved concentration versus time using SEAM3D-PUP and five TSCF values for the closed-system model depicted in Figure 3.1.
TSCF = 1.0 TSCF = 0.75 TSCF = 0.5 TSCF = 0.25 TSCF = 0.0 Mass Conc. Mass Conc. Mass Conc. Mass Conc. Mass Conc.
Time (d)
(g) (mg/L) (g) (mg/L) (g) (mg/L) (g) (mg/L) (g) (mg/L) 0 0 10 0 10 0 10 0 10 0.0 10.0 2 2000 9.95 1500 9.9625 1000 9.975 500 9.9875 0.0 10.0 4 3990 9.9002 2994 9.9251 1997 9.9501 999 9.975 0.0 10.0 6 5970 9.8507 4483 9.8879 2992 9.9252 1498 9.9625 0.0 10.0 8 7940 9.8015 5966 9.8508 3985 9.9004 1996 9.9501 0.0 10.0 10 9900 9.7525 7444 9.8139 4975 9.8756 2493 9.9377 0.0 10.0
Table 4.3. Comparison of concentration and mass removed through direct uptake versus time using SEAM3D-PUP to results using SEAM3D-SSM for the closed-system, two stress period model – case (3.1.2).
SEAM3D-PUP SEAM3D-SSM Time (d) Conc. (mg/L) Mass Out (g) Conc. (mg/L) Mass Out (g)0 10 0 10 0 2 9.95 2000 9.95 2000 4 9.9002 3990 9.900249 3990 6 9.8507 5970.1 9.850748 5970.1 8 9.8015 7940.2 9.801495 7940.2 10 9.7525 9900.5 9.752487 9900.5 12 9.7525 9900.5 9.752487 9900.5 14 9.7525 9900.5 9.752487 9900.5 16 9.7525 9900.5 9.752487 9900.5 18 9.7525 9900.5 9.752487 9900.5 20 9.7525 9900.5 9.752487 9900.5
83
Table 4.4. Simulation results for mass removed by direct uptake and dissolved concentration versus time using SEAM3D-PUP and five TSCF values for the closed-system model, four stress period model.
TSCF = 0.0 TSCF = 0.25 TSCF = 0.5 TSCF = 0.75 TSCF = 1.0 Time (d) Conc.
mg/L Mass
g Conc. mg/L
Mass g
Conc. mg/L
Mass g
Conc. mg/L
Mass g
Conc. mg/L
Mass g
0 10 0 10.00 0 10.00 0 10.00 0 10.00 0 2 10 0 9.99 500 9.975 1000 9.9625 1500 9.95 2000 4 10 0 9.98 999.37 9.9501 1997.5 9.9251 2994.4 9.9002 3990 6 10 0 9.96 1498.1 9.9252 2992.5 9.8879 4483.1 9.8507 5970.1 8 10 0 9.95 1996.3 9.9004 3985 9.8508 5966.3 9.8015 7940.2 10 10 0 9.94 2493.8 9.8756 4975.1 9.8139 7444 9.7525 9900.5 12 10 0 9.94 2493.8 9.8756 4975.1 9.8139 7444 9.7525 9900.5 14 10 0 9.94 2493.8 9.8756 4975.1 9.8139 7444 9.7525 9900.5 16 10 0 9.94 2493.8 9.8756 4975.1 9.8139 7444 9.7525 9900.5 18 10 0 9.94 2493.8 9.8756 4975.1 9.8139 7444 9.7525 9900.5 20 10 0 9.94 2493.8 9.8756 4975.1 9.8139 7444 9.7525 9900.5 22 10 0 9.93 2990.6 9.8509 5962.6 9.7771 8916.1 9.7037 11851 24 10 0 9.91 3486.9 9.8263 6947.7 9.7404 10383 9.6552 13792 26 10 0 9.90 3982.5 9.8017 7930.3 9.7039 11844 9.6069 15723 28 10 0 9.89 4477.6 9.7772 8910.5 9.6675 13299 9.5589 17644 30 10 0 9.88 4972 9.7528 9888.2 9.6313 14749 9.5111 19556 32 10 0 9.88 4972 9.7528 9888.2 9.6313 14749 9.5111 19556 34 10 0 9.88 4972 9.7528 9888.2 9.6313 14749 9.5111 19556 36 10 0 9.88 4972 9.7528 9888.2 9.6313 14749 9.5111 19556 38 10 0 9.88 4972 9.7528 9888.2 9.6313 14749 9.5111 19556 40 10 0 9.88 4972 9.7528 9888.2 9.6313 14749 9.5111 19556
84
Table 4.5. Simulation results for mass removed by direct uptake for TSCF = 1.0 using SEAM3D-SSM and SEAM3D-PUP for the model shown in Figure 3.6.
TIME TOTAL IN TOTAL OUT SOURCES SINKS NET MASS
FROM FLUID-STORAGE
TOTAL MASS IN AQUIFER
DISCREPANCY (%)
d g g g g g g TOTAL IN-OUT ALTERNATIVE
182.5 5.12E+05 -5.12E+05 5.12E+05 -4.023E-05 0 5.12E+05 1.83E-05 2.01E-04 365 8.84E+05 -8.84E+05 8.84E+05 -111.03 0 8.84E+05 2.12E-05 8.82E-05 547.5 1.25E+06 -1.25E+06 1.25E+06 -7225.9 0 1.24E+06 5.00E-05 8.90E-05 730 1.62E+06 -1.62E+06 1.62E+06 -55039 0 1.56E+06 3.09E-05 6.77E-05 912.5 1.98E+06 -1.98E+06 1.98E+06 -180154 0 1.80E+06 4.42E-05 -7.33E-05 1095 2.35E+06 -2.35E+06 2.35E+06 -387812 0 1.96E+06 5.33E-05 -6.79E-05 1277.5 2.71E+06 -2.71E+06 2.71E+06 -659936 0 2.05E+06 1.57E-04 1.20E-04 1460 3.08E+06 -3.08E+06 3.08E+06 -973960 0 2.10E+06 6.50E-05 5.28E-05 1642.5 3.44E+06 -3.44E+06 3.44E+06 -1312250 0 2.13E+06 -4.36E-05 -7.99E-05 1825 3.81E+06 -3.81E+06 3.81E+06 -1663540 0 2.15E+06 -1.05E-04 -1.64E-04 2007.5 4.17E+06 -4.17E+06 4.17E+06 -2021460 0 2.15E+06 -1.50E-04 -2.13E-04 2190 4.54E+06 -4.54E+06 4.54E+06 -2382630 0 2.16E+06 9.91E-05 5.51E-06 2372.5 4.91E+06 -4.91E+06 4.91E+06 -2745350 0 2.16E+06 2.96E-04 2.55E-04 2555 5.27E+06 -5.27E+06 5.27E+06 -3108810 0 2.16E+06 4.93E-04 4.60E-04 2737.5 5.64E+06 -5.64E+06 5.64E+06 -3472630 0 2.16E+06 6.65E-04 6.25E-04 2920 6.00E+06 -6.00E+06 6.00E+06 -3836620 0 2.17E+06 8.25E-04 8.33E-04 3102.5 6.37E+06 -6.37E+06 6.37E+06 -4200700 0 2.17E+06 9.82E-04 9.89E-04 3285 6.73E+06 -6.73E+06 6.73E+06 -4564840 0 2.17E+06 1.10E-03 1.09E-03 3467.5 7.10E+06 -7.10E+06 7.10E+06 -4929010 0 2.17E+06 1.25E-03 1.25E-03
SEA
M3D
-SSM
3650 7.46E+06 -7.46E+06 7.46E+06 -5293210 0 2.17E+06 1.31E-03 1.30E-03
182.5 5.12E+05 -5.12E+05 5.12E+05 -4.02275E-05 0 5.12E+05 2.44E-05 2.01E-04
365 8.84E+05 -8.84E+05 8.84E+05 -111.03 0 8.84E+05 2.83E-05 8.11E-05
547.5 1.25E+06 -1.25E+06 1.25E+06 -7225.9 0 1.24E+06 5.00E-05 8.90E-05
730 1.62E+06 -1.62E+06 1.62E+06 -55039 0 1.56E+06 3.09E-05 6.02E-05
912.5 1.98E+06 -1.98E+06 1.98E+06 -180154 0 1.80E+06 5.68E-05 -7.25E-05
1095 2.35E+06 -2.35E+06 2.35E+06 -387812 0 1.96E+06 6.39E-05 -6.79E-05
1277.5 2.71E+06 -2.71E+06 2.71E+06 -659936 0 2.05E+06 1.57E-04 1.20E-04
1460 3.08E+06 -3.08E+06 3.08E+06 -973960 0 2.10E+06 7.31E-05 5.28E-05
1642.5 3.44E+06 -3.44E+06 3.44E+06 -1312250 0 2.13E+06 -4.36E-05 -7.99E-05
1825 3.81E+06 -3.81E+06 3.81E+06 -1663540 0 2.15E+06 -1.05E-04 -1.64E-04
2007.5 4.17E+06 -4.17E+06 4.17E+06 -2021460 0 2.15E+06 -1.56E-04 -2.13E-04
2190 4.54E+06 -4.54E+06 4.54E+06 -2382630 0 2.16E+06 8.81E-05 1.10E-05
2372.5 4.91E+06 -4.91E+06 4.91E+06 -2745350 0 2.16E+06 2.96E-04 2.55E-04
2555 5.27E+06 -5.27E+06 5.27E+06 -3108810 0 2.16E+06 4.84E-04 4.65E-04
2737.5 5.64E+06 -5.64E+06 5.64E+06 -3472630 0 2.16E+06 6.74E-04 6.25E-04
2920 6.00E+06 -6.00E+06 6.00E+06 -3836620 0 2.17E+06 8.41E-04 8.37E-04
3102.5 6.37E+06 -6.37E+06 6.37E+06 -4200700 0 2.17E+06 9.82E-04 9.93E-04
3285 6.73E+06 -6.73E+06 6.73E+06 -4564840 0 2.17E+06 1.11E-03 1.10E-03
3467.5 7.10E+06 -7.10E+06 7.10E+06 -4929010 0 2.17E+06 1.26E-03 1.26E-03
SEA
M3D
-PU
P
3650 7.46E+06 -7.46E+06 7.46E+06 -5293210 0 2.17E+06 1.33E-03 1.32E-03
85
Table 4.6. Concentration results for the three observation points along ET zone for both
SEAM3D-SSM and SEAM3D-PUP for TSCF = 1.0 – case study (4.1.3). SEAM3D-SSM SEAM3D-PUP Time, d
j = 45 50 56 45 50 56 182.5 1.2E-05 8.532E-09 6.972E-19 1.2E-05 8.53E-09 6.97E-19 365 0.927818 0.1377061 9.643E-07 0.92782 0.13771 9.64E-07 547.5 14.17174 5.6473708 0.0012467 14.172 5.6474 1.25E-03 730 41.40879 25.322618 0.0329837 41.409 25.323 3.30E-02 912.5 66.84296 51.114315 0.2064269 66.843 51.114 0.20643 1095 83.43466 72.227707 0.6465797 83.435 72.228 0.64658 1277.5 92.44934 85.772156 1.3849308 92.449 85.772 1.3849 1460 96.87458 93.346741 2.3662646 96.875 93.347 2.3663 1642.5 98.92119 97.247032 3.5080974 98.921 97.247 3.5081 1825 99.83408 99.152687 4.7399039 99.834 99.153 4.7399 2007.5 100.232 100.05186 6.0138955 100.23 100.05 6.0139 2190 100.4034 100.46619 7.3017144 100.4 100.47 7.3017 2372.5 100.4764 100.65306 8.5879049 100.48 100.65 8.5879 2555 100.5074 100.73701 9.8645144 100.51 100.74 9.8645 2737.5 100.5204 100.77406 11.12765 100.52 100.77 11.128 2920 100.5259 100.79036 12.375437 100.53 100.79 12.375 3102.5 100.5282 100.79752 13.607041 100.53 100.8 13.607 3285 100.5291 100.80064 14.822114 100.53 100.8 14.822 3467.5 100.5295 100.80196 16.020529 100.53 100.8 16.021 3650 100.5296 100.8025 17.202309 100.53 100.8 17.202
86
Table 4.7. Simulation results for dissolved concentration versus time using SEAM3D-PUP
and five TSCF values and compared to SEAM3D-SSM for the observation point (24, 50, 1) for the case study (4.1.3).
Conc. Mg/L SEAM3D-PUP Time d TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.0
SEAM3D-SSM
182.5 8.53E-09 8.63E-09 8.74E-09 8.84E-09 8.95E-09 8.53E-09 365 0.13771 0.14465 0.15203 0.15988 0.16824 0.137706 547.5 5.6474 6.1159 6.6375 7.2201 7.8726 5.647371 730 25.323 28.228 31.627 35.634 40.393 25.32262 912.5 51.114 58.524 67.652 79.043 93.559 51.11432 1095 72.228 84.677 100.82 122.34 151.87 72.22771 1277.5 85.772 102.55 125.45 158.05 206.79 85.77216 1460 93.347 113.33 141.92 185.44 256.77 93.34674 1642.5 97.247 119.39 152.43 206.11 302.96 97.24703 1825 99.153 122.67 159.03 221.83 346.78 99.15269 2007.5 100.05 124.4 163.15 233.95 389.24 100.0519 2190 100.47 125.3 165.74 243.4 430.92 100.4662 2372.5 100.65 125.77 167.37 250.84 472.13 100.6531 2555 100.74 126.02 168.4 256.73 513.02 100.737 2737.5 100.77 126.14 169.06 261.4 553.65 100.7741 2920 100.79 126.21 169.47 265.12 594.07 100.7904 3102.5 100.8 126.24 169.74 268.08 634.28 100.7975 3285 100.8 126.26 169.91 270.44 674.29 100.8006 3467.5 100.8 126.27 170.02 272.31 714.11 100.802 3650 100.8 126.27 170.09 273.81 753.73 100.8025 Table 4.8 Model parameters for the flow and transport with root sorption case study (4.2.1)
SEAM3D-RCT (in GMS) SEAM3D-PUP
ET package QET = 0.0 QET = 0.0 RCT package Kd = 14.3×10-8, 7.143×10-8, and 5.0×10-8 Kd = 0.0
RCF Package N/A RCF =14.3×10-8, 7.143×10-8, and 5.0×10-8 for all model domain.
87
Table 4.9. SEAM3D-PUP and SEAM3D-RCT results for dissolved concentration at the
observation point (24, 50, 1) for the flow and transport with root sorption case study – f = 1.0.
Concentration, SEAM3D-PUP Concentration, SEAM3D-RCT Time, d RCF= 0.0
(R=1.0) 7.14×10-8 (R=1.50)
14.3×10-8 (R=2.0) Kd = 0.0
(R=1.0) 7.14×10-8 (R=1.50)
14.3×10-8 (R=2.0)
182.5 2.51E-09 2.95E-22 0 2.51E-09 2.95E-22 0 365 8.68E-02 8.02E-05 5.73E-09 0.086838 8.02E-05 5.73E-09 547.5 4.3175 9.39E-02 1.16E-03 4.317454 0.093881 0.001165 730 21.417 1.8098 9.99E-02 21.41726 1.809845 0.09986 912.5 46.109 8.6658 1.0713 46.10904 8.665835 1.071324 1095 67.815 21.607 4.5137 67.81548 21.60671 4.513667 1277.5 82.59 37.92 11.517 82.58997 37.91979 11.51749 1460 91.177 54.046 21.722 91.177 54.04603 21.72207 1642.5 95.77 67.791 33.796 95.76965 67.79108 33.79574 1825 98.036 78.338 46.169 98.03575 78.33845 46.16918 2007.5 99.144 85.921 57.691 99.14414 85.92098 57.69147 2190 99.635 91.071 67.653 99.63529 91.07111 67.65306 2372.5 99.878 94.47 75.849 99.878 94.47012 75.84943 2555 99.965 96.624 82.301 99.96533 96.6244 82.30144 2737.5 100.02 97.984 87.247 100.021 97.98402 87.24729 2920 100.03 98.805 90.925 100.0264 98.80501 90.92463 3102.5 100.04 99.314 93.625 100.0442 99.31432 93.62485 3285 100.03 99.607 95.558 100.0345 99.60679 95.55785 3467.5 100.04 99.791 96.94 100.0445 99.79102 96.94 3650 100.03 99.888 97.902 100.0329 99.88799 97.9018
88
Table 4.10. SEAM3D-PUP and SEAM3D-RCT results for mass removal at the observation
point (24, 50, 1) for the flow and transport with root sorption case study (4.2.1) – f = 1.0. Mass Sinks by Sorption, g
R = 1.0 R = 1.50 R = 2.0 Time d
SEAM3D/PUP SEAM3D/RCT SEAM3D/PUP SEAM3D/RCT SEAM3D/PUP SEAM3D/RCT 182.5 0 0 0 0 0 0 365 -5.34E-20 -5.33545E-20 0 0 0 0 547.5 -1.73E-06 -1.72796E-06 -1.25787E-19 -1.25787E-19 -6.74083E-37 -6.74081E-37 730 -0.07007 -0.0700696 -9.49089E-09 -9.4909E-09 -2.74002E-19 -2.74002E-19 912.5 -19.388 -19.388 -0.000294574 -0.000294574 -4.1154E-10 -4.1154E-10 1095 -606.58 -606.58 -0.12633 -0.12633 -6.78682E-06 -6.78682E-06 1277.5 -5903.6 -5903.6 -6.9958 -6.9958 -0.00291474 -0.00291474 1460 -28593 -28593 -119.72 -119.72 -0.19725 -0.19725 1642.5 -88591 -88591 -973.8 -973.8 -4.4008 -4.4008 1825 -203531 -203531 -4794 -4794 -47.357 -47.357 2007.5 -380364 -380364 -16552 -16552 -306.58 -306.58 2190 -614512 -614512 -44105 -44105 -1371.8 -1371.8 2372.5 -894531 -894531 -96787 -96787 -4652.8 -4652.8 2555 -1207350 -1207350 -183131 -183131 -12752 -12752 2737.5 -1541710 -1541710 -308882 -308882 -29570 -29570 2920 -1889250 -1889250 -476080 -476080 -60016 -60016 3102.5 -2244450 -2244450 -683255 -683255 -109383 -109383 3285 -2603890 -2603890 -926350 -926350 -182601 -182601 3467.5 -2965640 -2965640 -1199890 -1199890 -283553 -283553 3650 -3328580 -3328580 -1498020 -1498020 -414681 -414681
89
Table 4.11. SEAM3D-PUP and SEAM3D-RCT results for dissolved concentration at the observation point (24, 50, 1) for the flow and transport with root sorption case study – f < 1.0, and f = 1.0.
SEAM3D-PUP SEAM3D-RCT RCF = 14.3×10-8 RCF = 14.3×10-8 Kd = 8.93×10-8 Kd = 14.3×10-8
Time Days
R = 2.0 (f = 1.0) Rav. = 1.62 (f < 1.0) R=1.625 R=2.0 182.5 0 1.69E-24 0 0 365 5.73E-09 6.39E-06 1.04639E-05 5.73E-09 547.5 1.16E-03 2.26E-02 0.032857418 0.0011648 730 9.99E-02 0.69298 0.903235376 0.09986043 912.5 1.0713 4.3541 5.281054497 1.07132351 1095 4.5137 13.024 15.02291298 4.51366711 1277.5 11.517 26.055 28.97231102 11.5174875 1460 21.722 40.934 44.29027176 21.7220707 1642.5 33.796 55.27 58.5772171 33.795742 1825 46.169 67.534 70.46270752 46.169178 2007.5 57.691 77.25 79.65240479 57.6914749 2190 67.653 84.479 86.34081268 67.6530609 2372.5 75.849 89.659 91.04132843 75.8494339 2555 82.301 93.225 94.21688843 82.3014374 2737.5 87.247 95.641 96.33469391 87.2472916 2920 90.925 97.224 97.69844818 90.9246292 3102.5 93.625 98.264 98.58332825 93.6248474 3285 95.558 98.919 99.13118744 95.5578537 3467.5 96.94 99.346 99.48442841 96.9400024 3650 97.902 99.603 99.6929245 97.9018021 Table 4.12. Model parameters for the flow and transport with root sorption case study (4.2.3).
SEAM3D-RCT (in GMS) SEAM3D-PUP
ET package QET = 0.0 QET = 0.0 RCT package Kd = 14.3×10-8 (Ro = 1.0) Kd = 7.143×10-8 (Ro = 0.5)
RCF Package N/A RCF = 7.143×10-8 for all model domain (Ro = 0.5)*.
* 0.1−= RRo
90
Table 4.13. SEAM3D-PUP and SEAM3D-RCT results for dissolved concentration at the observation point (24, 50, 1) for the flow and transport with root sorption where 50% of the retardation factor is due to root sorption, and 50% is due to soil sorption – f = 1.0.
SEAM3D-PUP SEAM3D-RCT
Time f = 1.0,
Kd = 7.143×10-8, RCF = 7.143×10-8
Kd = 14.3×10-8 Kd = 7.143×10-
8
R=2.0 R=2.0 R=1.5 182.5 0 0 2.95E-22 365 5.97E-09 5.73E-09 8.01901E-05 547.5 1.18E-03 0.0011648 0.093880534 730 0.10072 0.09986043 1.809844971 912.5 1.078 1.071323514 8.665835381 1095 4.535 4.513667107 21.60671043 1277.5 11.56 11.51748753 37.91978836 1460 21.785 21.72207069 54.04603195 1642.5 33.874 33.79574203 67.79107666 1825 46.254 46.16917801 78.33844757 2007.5 57.775 57.69147491 85.92098236 2190 67.73 67.65306091 91.07110596 2372.5 75.917 75.8494339 94.47012329 2555 82.358 82.30143738 96.62440491 2737.5 87.293 87.24729156 97.98402405 2920 90.96 90.92462921 98.80500793 3102.5 93.652 93.62484741 99.3143158 3285 95.579 95.5578537 99.60678864 3467.5 96.956 96.94000244 99.79102325 3650 97.913 97.90180206 99.88798523
91
Table 4.14. Concentration versus time for the three middle observation points (using SEAM3D-PUP and SEAM3D-RCT) for the model in Figure 4.6, with root sorption in ET area only for f = 1.0 (case study 4.2.3.1).
Concentration, SEAM3D-PUP Concentration, SEAM3D-RCT Cell # Cell # Time
d 1,24,45 1,24,50 1,24,56 1,24,45 1,24,50 1,24,56
182.5 7.17E-15 7.01E-23 0 7.18E-15 2.95E-22 0 365 2.77E-03 3.88E-05 6.09E-08 0.002836 8.02E-05 3.77E-07 547.5 0.5472 6.05E-02 2.64E-03 0.567037 0.093881 0.006977 730 5.1989 1.3229 0.18961 5.402804 1.809845 0.367179 912.5 16.981 6.8422 1.8616 17.62783 8.665835 3.014644 1095 33.551 18.019 7.3016 34.71582 21.60671 10.48911 1277.5 50.662 32.968 17.442 52.19154 37.91979 22.97352 1460 65.367 48.556 30.93 67.01652 54.04603 38.1467 1642.5 76.701 62.514 45.454 78.26171 67.79108 53.2793 1825 84.791 73.753 58.961 86.13787 78.33845 66.41999 2007.5 90.309 82.209 70.363 91.39576 85.92098 76.8526 2190 93.925 88.228 79.274 94.75832 91.07111 84.54449 2372.5 96.251 92.377 85.911 96.86582 94.47012 89.97752 2555 97.707 95.135 90.62 98.14613 96.6244 93.62997 2737.5 98.617 96.949 93.89 98.9231 97.98402 96.04946 2920 99.169 98.103 96.072 99.37786 98.80501 97.57979 3102.5 99.51 98.844 97.529 99.65079 99.31432 98.56155 3285 99.711 99.299 98.458 99.80378 99.60679 99.15042 3467.5 99.836 99.59 99.068 99.89709 99.79102 99.52755 3650 99.906 99.76 99.439 99.94511 99.88799 99.73798
92
Table 4.15. Results for mass removal by direct uptake and root sorption versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.1.
No Sorption With Sorption, RCF=Kd=7.14E-8 PUP PUP Time
d TSCF=1.0 TSCF=0.5
SSM TSCF=1.0 TSCF=0.50
SSM+RCTS
182.5 4.02E-05 2.02E-05 4.02E-05 4.445E-15 2.2236E-15 5.88941E-15365 111.03 58.211 111.03 0.19045 0.0966125 0.2422 547.5 7225.9 3999.3 7225.9 141.88 73.598 178.75 730 55039 32386 55039 3224 1719 4016.2 912.5 180154 112843 180154 20444 11247 25117 1095 387812 257841 387812 69213 39380 83717 1277.5 659936 463210 659936 163878 96557 195010 1460 973960 716844 973960 310368 189418 363359 1642.5 1312250 1005750 1312250 506753 320161 584044 1825 1663540 1319100 1663540 746376 487571 847695 2007.5 2021460 1648860 2021460 1020820 688359 1143940 2190 2382630 1989410 2382630 1321880 918337 1463580 2372.5 2745350 2337000 2745350 1642530 1173090 1799270 2555 3108810 2689160 3108810 1977190 1448530 2145670 2737.5 3472630 3044270 3472630 2321650 1740970 2498990 2920 3836620 3401320 3836620 2672850 2047300 2856740 3102.5 4200700 3759640 4200700 3028630 2364870 3217240 3285 4564840 4118800 4564840 3387480 2691530 3579460 3467.5 4929010 4478540 4929010 3748420 3025520 3942750 3650 5293210 4838680 5293210 4110710 3365390 4306690
93
Table 4.16. Results for dissolved concentration versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.1.
No Sorption With Sorption, RCF=Kd=7.14E-8 PUP PUP Time
d TSCF=1.0 TSCF=0.5
SSM TSCF=1.0 TSCF=0.50
SSM+RCT
182.5 8.53E-09 8.74E-09 8.53E-09 2.42E-22 2.42E-22 1.02E-21 365 0.13771 0.15203 0.137706 7.88E-05 8.25E-05 0.00017114 547.5 5.6474 6.6375 5.647371 9.26E-02 0.10063 0.147660106730 25.323 31.627 25.32262 1.7828 2.0105 2.477855444912.5 51.114 67.652 51.11432 8.494 9.9313 10.851209641095 72.228 100.82 72.22771 21.118 25.598 25.437255861277.5 85.772 125.45 85.77216 37.007 46.472 42.665454861460 93.347 141.92 93.34674 52.755 68.574 58.802448271642.5 97.247 152.43 97.24703 66.271 89.014 71.953659061825 99.153 159.03 99.15269 76.775 106.35 81.6647644 2007.5 100.05 163.15 100.0519 84.468 120.38 88.433692932190 100.47 165.74 100.4662 89.831 131.37 92.903442382372.5 100.65 167.37 100.6531 93.486 139.87 95.7955246 2555 100.74 168.4 100.737 95.899 146.39 97.587402342737.5 100.77 169.06 100.7741 97.493 151.42 98.707794192920 100.79 169.47 100.7904 98.512 155.29 99.367782593102.5 100.8 169.74 100.7975 99.18 158.29 99.778961183285 100.8 169.91 100.8006 99.594 160.63 100.00501253467.5 100.8 170.02 100.802 99.869 162.47 100.15198523650 100.8 170.09 100.8025 100.03 163.91 100.221077
94
Table 4.17. Dissolved concentration results for SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.2.
Concentration, mg/L Time d GMS PUP 182.5 0 0.00E+00 365 1.98E-08 2.04E-08 547.5 0.00221116 2.23E-03 730 0.15716679 0.15828 912.5 1.50656378 1.5143 1095 5.88046932 5.9035 1277.5 14.1743917 14.218 1460 25.5782661 25.639 1642.5 38.4262619 38.498 1825 51.0445251 51.119 2007.5 62.3712158 62.442 2190 71.8477325 71.911 2372.5 79.4293365 79.483 2555 85.2455826 85.289 2737.5 89.6107025 89.645 2920 92.7893448 92.816 3102.5 95.0880737 95.108 3285 96.7042007 96.719 3467.5 97.848732 97.86 3650 98.6308365 98.639
95
Table 4.18. Results of mass removal by direct uptake and root sorption using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages (GMS) for case study 4.3.2.
TIME TOTAL IN TOTAL OUT SOURCES SINKS NET MASS
FROM FLUID-STORAGE
TOTAL MASS IN AQUIFER
DISCREPANCY (%)
d g g g g g g TOTAL IN-OUT ALTERNATIVE
182.5 638789 -638789 638784 0 0 638784 0.00001957 -0.000009784365 1028980 -1028980 1028960 -0.00016 -35.75 1028930 0.00001822 -0.00001216
547.5 1403180 -1403180 1403160 -3.5726 15.406 1403170 0.00005345 0.00004159730 1772400 -1772400 1772350 -254.84 -63.969 1772030 0.00007053 -0.00004894
912.5 2139800 -2139800 2139730 -3018.5 14.656 2136730 0.00005842 -0.000020451095 2506170 -2506170 2506070 -15314 -84.594 2490670 0.00008978 0.0001057
1277.5 2872200 -2872200 2872070 -48285 9.0313 2823800 0.00007834 0.000095481460 3237920 -3237920 3237760 -113223 -99.469 3124430 0.00007721 0.0001803
1642.5 3603610 -3603610 3603410 -217983 -1.8438 3385420 0.00007631 0.00017431825 3969170 -3969170 3968930 -365627 -116.47 3603180 0.00008818 0.0002528
2007.5 4334780 -4334770 4334490 -554961 -18.969 3779500 0.00005767 0.00015932190 4700310 -4700310 4699980 -781930 -133.72 3917910 0.00003191 0.00007115
2372.5 5065920 -5065920 5065530 -1041010 -38.219 4024480 0.00002961 0.000070952555 5431460 -5431460 5431020 -1326320 -154.72 4104540 0.00002762 -0.00005466
2737.5 5797070 -5797070 5796560 -1632380 -64.969 4164120 0.0000345 -0.000070622920 6162630 -6162630 6162060 -1954390 -166.59 4207500 0.00006491 -0.00004615
3102.5 6528270 -6528260 6527610 -2288410 -82.594 4239120 0.00009191 -0.0000090963285 6893840 -6893830 6893110 -2631300 -173.34 4261630 0.00007253 0.00007843
3467.5 7259500 -7259490 7258670 -2980670 -95.844 4277900 0.00007576 0.00006415
SEA
M3D
-PU
P
3650 7625090 -7625080 7624190 -3334690 -183.09 4289310 0.00007869 0.00007747182.5 638931 -638930 638931 0 0 638930 0.00001956 0.0001125
365 1029150 -1029150 1029130 -0.000154673 -35.562 1029090 0.00001215 0.00008197
547.5 1403370 -1403370 1403330 -3.5388 16.875 1403340 0.00007126 0.0001843
730 1772610 -1772610 1772530 -253.14 -60.25 1772220 0.00006347 0.00003466
912.5 2140020 -2140020 2139910 -3002.8 17.062 2136930 0.00003505 -0.000004427
1095 2506400 -2506400 2506260 -15250 -83.688 2490920 0.00001995 0.00009036
1277.5 2872450 -2872450 2872260 -48115 7.1875 2824150 0.00002611 -0.00001618
1460 3238180 -3238180 3237940 -112885 -105.31 3124950 0.00002316 0.00004271
1642.5 3603880 -3603880 3603600 -217425 -6.4375 3386170 0.00001387 0.00001778
1825 3969460 -3969460 3969120 -364822 -123.81 3604180 0 0.000008661
2007.5 4335080 -4335080 4334690 -553902 -25.062 3780760 0 -0.00004326
2190 4700620 -4700620 4700170 -780619 -142.44 3919420 0.00002127 -0.00006383
2372.5 5066220 -5066220 5065710 -1039460 -47.562 4026200 -0.000009869 -0.00002714
2555 5431780 -5431780 5431210 -1324590 -156.06 4106470 -0.00002762 -0.0001116
2737.5 5797410 -5797410 5796760 -1630490 -64.438 4166210 -0.00001725 -0.0001456
2920 6162970 -6162970 6162240 -1952360 -169.19 4209720 0 -0.0000781
3102.5 6528620 -6528620 6527800 -2286280 -80.938 4241450 0.00001532 -0.00006415
3285 6894210 -6894200 6893310 -2629090 -177.69 4264040 0.0000145 0.00003355
3467.5 7259870 -7259870 7258860 -2978390 -94.438 4280380 0.00003444 0.00003186
GM
S\SE
AM
3D
3650 7625480 -7625470 7624390 -3332370 -184.19 4291830 0.00005246 0.0001189
96
Recharge ET
10 m
h = 8m0
Figure 4.1. Schematic of a closed system model for testing the direct uptake feature using
the SEAM3D-RDP.
97
9.7
9.75
9.8
9.85
9.9
9.95
10
0 2 4 6 8 10Time (d)
Solu
te C
once
ntra
tion
(mg/
L)
0
2000
4000
6000
8000
10000
12000
Mas
s O
ut (g
)
Conc. (SEAM3D-SSM) Conc. (SEAM3D-PUP)Mass (SEAM3D-SSM) Mass (SEAM3D-PUP)
Figure 4.2. Simulated dissolved concentration and mass removed by direct uptake versus
time from SEAM3D-PUP and SEAM3D-SSM with TSCF = 1.0 for the closed-system, single
stress period model in Figure 3.1.
98
9.7
9.75
9.8
9.85
9.9
9.95
10
0 1 2 3 4 5 6 7 8 9 10
Time, d
Dis
solv
ed C
once
ntra
tion
(mg/
L)
TSCF=1.0 0.75 0.5 0.25 0 GMS/SEAM3D
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
0 1 2 3 4 5 6 7 8 9 10
Time, d
Mas
s O
ut (g
)
TSCF = 1.0 0.75 0.5 0.25 0 GMS/SEAM3D
Figure 4.3. Simulated dissolved concentration (top) and mass removed by direct uptake
(bottom) versus time using SEAM3D-PUP for the closed-system model in Figure 3.1 for the
range of TSCF values, varying from 0 to 1.0.
99
ET R
ate
Time10 20
0.01
9.7
9.75
9.8
9.85
9.9
9.95
10
0 10
Time (d)
Dis
solv
ed C
once
ntra
tion
(mg/
L)
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Mas
s O
ut (g
)
20
Conc. (SEAM3D-PUP) Conc. (SEAM3D-SSM)Mass (SEAM3D-PUP) Mass (SEAM3D-SSM)
Figure 4.4. Simulated dissolved concentration and mass removed by direct uptake versus
time from SEAM3D-PUP and SEAM3D-SSM with TSCF = 1.0 for the closed-system model
in Figure 4.1 with two stress periods with variable rates of evapotranspiration (top).
100
9.4
9.5
9.6
9.7
9.8
9.9
10
0 10 20 30 4
Time (d)
Dis
solv
ed C
once
ntra
tion
(mg/
L)
0
SEAM3D/PUP, TSCF=0 SEAM3D/PUP, TSCF=0.25 SEAM3D/PUP, TSCF=0.50SEAM3D/PUP, TSCF=0.75 SEAM3D/PUP, TSCF=1.0 SEAM3D/SSM, TSCF=1.0
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
0 10 20 30
Time (d)
Mas
s O
ut (g
)
40
SEAM3D/PUP, TSCF=0 SEAM3D/PUP, TSCF=0.25 SEAM3D/PUP, TSCF=0.50SEAM3D/PUP, TSCF=0.75 SEAM3D/PUP, TSCF=1.0 SEAM3D/SSM, TSCF=1.0
Figure 4.5. Simulated dissolved concentration (top) and mass removed by direct uptake
(bottom) versus time using SEAM3D-PUP for a four stress period, closed-system model in
Figure 3.1 for the range of TSCF values, varying from 0 to 1.0.
101
200.0
100.
0
Q = 0.4 m /d (each)in
Con
stan
t Hea
d
90.0 20.0 90.0
ET
(10.0)
(0.0)
(8.0)
i=50
j=100
ETLand Surface
Impermeable
3
Conc. = 100 mg /L (each)
Obs. Points
Figure 4.6. Conceptual model for case study 3.1.3, flow and transport with direct uptake in
the ET area (no root sorption; TSCF is T, and RCF is F). Three observation points are
noted: (i, j, k) = (24, 45, 1), (24, 50, 1), and (24, 56, 1).
102
0
1000
2000
3000
4000
5000
6000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s O
ut (E
T), g
TSCF=1.0 0.75 0.50 0.25 SEAM3D-SSM
Figure 4.7. Mass removal by direct uptake versus time using SEAM3D-PUP and SEAM3D-
SSM for a one-stress period, flow-system model shown in Figure 4.6 for the range of TSCF
values, varying from 0.0 to 1.0.
103
0
10
20
30
40
50
60
70
80
90
100
110
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, Days
Con
c., m
g/L
SEAM3D-SSM 1, 24, 45 1, 24, 50 1, 24, 56SEAM3D-PUP 1,24,45 1,24,50 1,24,56
0
100
200
300
400
500
600
700
800
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Times, d
Con
c., m
g/L
TSCF=1.0 0.75 0.5 0.25 0.0 SEAM3D-SSM
Figure 4.8. Concentration versus time using SEAM3D-PUP and SEAM3D-SSM for a one-
stress period, flow-system model shown in Figure 4.6 (test case 4.1.3) for the three
observation points (top), and for the middle observation point for the range of TSCF values,
varying from 0.0 to 1.0 (bottom).
104
0
10
20
30
40
50
60
70
80
90
100
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, Days
Con
c., m
g/L
RCF=0.0 (R=1.0) RCF=5.0e-8, (R=1.35) RCF=14.3e-8, (R=2.0)
0
10
20
30
40
50
60
70
80
90
100
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, Days
Con
c., m
g/L
SEAM3D-PUP RCF=0.0, (R=1.0) RCF=5.0e-8, (R=1.35) RCF=14.3e-8, (R=2.0)
SEAM3D-RCT, Kd=0.0, (R=1.0) Kd=5.0e-8, (R=1.35) Kd=14.3e-8, (R=2.0)
Figure 4.9. Concentration versus time for the middle observation point, (i, j, k) = (24, 50, 1)
using SEAM3D-PUP (top), and comparing it with SEAM3D-RCT (bottom) for case study
4.2.1.
105
0
500
1000
1500
2000
2500
3000
3500
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s Si
nk, g
SEAM3D/PUP, R=1.5 SEAM3D/RCT, R=1.5 SEAM3D/PUP, R=2.0
SEAM3D/RCT, R=2.0 SEAM3D/PUP, R=1.0 SEAM3D/RCT, R=1.0
Figure 4.10. Mass removal versus time for the middle observation point, (i, j, k) = (24, 50, 1)
using SEAM3D-PUP and comparing it with SEAM3D-RCT for case study 4.2.1.
106
j=24 (t=3650)
7
7.5
8
8.5
9
9.5
10
0 20 40 60 80 1
cell #
Hyd
arul
ic H
ead,
m
00
0
10
20
30
40
50
60
70
80
90
100
0 356 712 1068 1424 1780 2136 2492 2848 3204 3560
Time , d
Con
c., m
g/L
SEAM3D-RCT, R=2.0 SEAM3D-PUP, f < 1.0, Rav.=1.62SEAM3D-RCT, R=1.625 SEAM3D-PUP, f=1.0, R=2.0
Figure 4.11. Hydraulic head distribution for r =24 (top), and concentration versus time for
the middle observation point, (i, j, k) = (24, 50, 1) using SEAM3D-PUP, and comparing it
with SEAM3D-RCT (bottom) for case study 4.2.2.
107
0
10
20
30
40
50
60
70
80
90
100
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
SEAM3D-RCT (Kd=7.143e-8), R=1.5
SEAM3D-PUP(soil:Kd=7.143e-8+Roots:RCF=7.143e-8), f=1.0, R=2.0
SEAM3D-RCT (Kd=14.3e-8), R=2.0
Figure 4.12. Concentration versus time for the middle observation point, (i, j, k) = (24, 50, 1)
using SEAM3D-PUP where 50% of the retardation is due to plant roots and 50% is due to
soil matrix, and comparing it with SEAM3D-RCT where 100% of the retardation is due to
soil matrix for case study 4.2.3.
108
0
10
20
30
40
50
60
70
80
90
100
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
PUP (j=24,i=45) PUP (j=24, i=50) PUP (i=24, j=56)W/out PUP (j=24, i=45) W/out PUP (i=24, j=50) W/out PUP (i=24, j=56)
Figure 4.13. Screen capture for the results of R in case study (4.2.3.1) showing R=2.0 in the
roots cells only, and R=1.5 everywhere else (top), and Concentration versus time for the three
middle observation points (Figure 4.6.), using SEAM3D-PUP and SEAM3D-RCT for case
study (4.2.3.1).
109
0
1000
2000
3000
4000
5000
6000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s, g
SEAM3D-PUP, TSCF=1.0, RCF=0.0 (R=1.0) SEAM3D-PUP, TSCF=0.5, RCF=0.0 (R=1.0)GMS/SEAM3D, Kd=0.0 (R=1.0) SEAM3D-PUP, TSCF=1.0, RCF=7.14e-8(R=1.5)SEAM3D-PUP, TSCF=0.5, RCF=7.14e-8(R=1.5) GMS/SEAM3D, Kd=7.14e-8 (R=1.5)
0
20
40
60
80
100
120
140
160
180
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
SEAM3D-PUP, TSCF=1.0, RCF=0.0 (R=1.0) SEAM3D-PUP, TSCF=0.5, RCF=0.0 (R=1.0)GMS/SEAM3D, Kd=0.0 (R=1.0) SEAM3D-PUP, TSCF=1.0, RCF=7.14e-8(R=1.5)SEAM3D-PUP, TSCF=0.5, RCF=7.14e-8(R=1.5) GMS/SEAM3D, Kd=7.14e-8 (R=1.5)
Figure 4.14. Mass removal by direct uptake and root sorption (top) and concentration
(bottom) versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages
for case study 4.3.1.
110
Qin
Con
stan
t Hea
d
90.0 20.0 90.0
RCF = 6.25×10-8
RC
F =
7.14
3×10
-8
RCF = 6.25×10-8
36
/10
75.1m
g R b
×=
ρ
00 m/ 10×=b 2. gρR 6 3 g2ρ =bR ×00. 10 6 / m3
R=2.0 R=2.0R=2.0K = 7.143×10d
-8
Figure 4.15. Conceptual model for case study 4.3.2, flow and transport with direct uptake in
the middle ET area and root sorption all over the model with different values of RCF.
111
0
10
20
30
40
50
60
70
80
90
100
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
GMS/SEAM3D SEAM3D/PUP
0
500
1000
1500
2000
2500
3000
3500
4000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s, g
SEAM3D-PUP GMS/SEAM3D
Figure 4.16. Concentration (top) and mass removal by direct uptake and root sorption
(bottom) versus time using SEAM3D-PUP and SEAM3D with the SSM and RCT Packages
for case study 4.3.2.
112
C h a p t e r 5
Simulation of a Phytoremediation System Using SEAM3D-PUP
5.1 Introduction
The remediation goals of each site may be different. For some sites, the remedial action objective
(RAO) is determined by a contaminant concentration downstream the source, and a remediation
approach is needed to shrink a plume to a certain length to avoid mixing with groundwater at a
withdrawal well or other receptor. For other cases, the RAO could be removing the solute mass from
the source as soon as possible and the plume length would not have much weight in the decision
making process. Furthermore, some remediation goals would be decreasing the mass-flux of the
groundwater carrying the contaminates to a certain level at a certain cross-section (normal to the flow
direction).
With these three different remediation goals in mind, the objective of the study in this chapter is to
examine the effect of a phytoremediation system with different geometric/hydrological arrangements
on:
1- The contaminant concentrations downstream the source (expressed in plume length at a
concentration 1% of the source concentration).
2- The solute mass removal from the aquifer
3- The mass-flux at different cross-sections downstream the contaminant source.
Towards achieving that goal, the plan of runs will include:
1- Effect of ET dimensions (width and length) on a dynamically steady state plume.
2- Effect of ET flux with respect to aquifer flux (UET/Uin).
3- Effect of a phytoremediation system when the contaminate source is removed on the
remediation outcomes.
This study demonstrates the usefulness of the SEAM3D-PUP package in addressing several issues
pertaining to the design or evaluation of a phytoremediation system that relies on phreatophytes.
113
Mass uptake or removing of contaminants is not explicitly addressed in most groundwater models
such as Evapotranspiration package in MODFLOW and the Source/Sink Mixing (SSM) package in
MT3DMS. Computational tools are needed to predict the effect of an engineered system of deep-
rooted poplars to provide a large degree of solute mass uptake, despite seasonal variation in water use
rates by the plantation. Modeling clearly has applications at phytoremediation sites for evaluating or
designing a remediation system with respect to factors such as tree planting density (Maximum ET
rate), plants dimensions of the phytoremediation system (WET and LET) relative to plume dimensions
(Lp), the contaminant source width (Ws), groundwater flow rate (Qin), and seasonal effects (represented
in different ET rates during stress periods).
The study in this chapter involves simulating a contaminant plume in a groundwater flow system in
an unconfined aquifer with constant flowrate under natural attenuation conditions and comparing the
results to the case of using the plants for controlling the plume dimensions and the mass flux
downstream the source. The natural attenuation processes include physical transport (advection,
dispersion) and/or biodegradation. The model will be used to determine the extents to which the
phytoremediation system will have on reducing the downstream plume concentration to a certain limit
at a specific distance, (Figure 5.1).
Source
In-Flow
ComplianceWell
Figure 5.1. The expected effect of using a phytoremediation system on reducing DS
concentration.
114
5.2 Model Description
In the first part of the study, a schematic of the model is designed to simulate the plume movement
under natural attenuation (NA) processes, and then with plant uptake, Figure 5.2. Sorption will have no
effect on the steady state plume except it will increase the time to reach the steady state. Although
biodegradation is enhanced due to the rhizosphere and can be simulated using the SEAM3D
Biodegradation Package, the rhizosphere effect will be directly considered.
The model dimensions are 1500m×500m, with uniform cell size of 5m×5m. The flowrate is kept
constant to the system at the left boundary by using injection wells along the whole length. The in-flux
= 1.5 m3/d/cell with a total inflow rate = 150 m3/day. Other model parameters were selected such that
to control the plume steady-state length to be within the two-thirds of total model length. The
longitudinal dispersivity was set to be equal to 10.0 d, and the ratio of transverse to longitudinal
dispersivity was equal to 0.20. The first order decay rate was set to be equal to 0.001 d-1. The previously
mentioned model parameters with initial concentration at the source cells equal to 1.0 mg/L, produced
a steady-state plume of approximately 940 m of length. The phytoremediation area was selected to be
1000m×300m which covers the steady-state plume, (Figure 5.2).
L
500.0
1500.0
Well Cells
ET Area
Constant-head CellsSource
Plume toe
Length of SS Plume, Lp
ET
ET
Figure 5.2. The conceptual model with the grid dimensions and boundary conditions.
The right boundary of the model is a constant-head boundary. The model will be run first without
the trees to estimate the time required for the plume to reach the steady state. Figure 5.3, which shows
that the mass-in reaches a constant value which indicates that the plume is stable after approximately
16 years (32 stress periods). The ET parameters such as ET surface elevation, hs, and extinction root
115
depth, d, were selected to maintain maximum ET rate in all the cells to be easy to compare with the
SEAM3D-PUP package. To compare the results from the SEAM3D-SSM runs using the ET package
versus the SEAM3D-PUP runs using the PUP package, TSCF is turned on, and set to a value equal to
1.0.
0
5000
10000
15000
20000
25000
30000
35000
40000
0 365730109514601825219025552920328536504015438047455110547558406205657069357300
Time, days
Mas
s-in
, g
GMS-NASEAM3D-PUP
Figure 5.3. Source mass in the system vs. time (using SEAM3D-SSM and SEAM3D-PUP)
under NA conditions.
After the plume reaches steady dimensions, the ET dimensions are first estimated to contain the
footprint of the stable plume of length =Lp under natural remediation conditions only, and then the
trees are introduced to the system. The ET rate is maximum at one stress period and then equals to
zero for the next stress period and the ET rate cycles between zero and maximum for the rest of
simulation periods which are equal to 20 stress periods of 182.5 days each (Figure 5.4). The maximum
ET rate is representing the plant capacity of evapotranspiration and can be controlled in the field by
controlling the number of trees in a unit area. A maximum ET rate of 0.001 (m3/d)/m2 per cell area
(25 m2) is equivalent to 6.6 gal/day/tree, if we use one tree in each cell (Table 2.2). The total QET is
kept at all runs equals to or less than the total inflow to prevent back water flow from the right
constant head boundary.
116
0 10 20t (years)
21
Time for the plume to be stable
ET
24
ET ET ET ET
22 23
star
ting
time
of th
e ne
w si
mul
atio
nt =
0
Figure 5.4. ET rate for different stress periods.
After one test case was performed, two different values for the ET lengths were used: LET=Lp and
LET=0.5Lp, five different values for ET width (relative to the contaminant source width are:
WET/Ws=3.0, 2.5, 2.0, 1.5, and 1.0), simultaneously with changing the TSCF values five times for each
model. Table 5.1 summaries the model parameters for different ET lengths, ET rates, and TSCF
values. Table 5.2 lists the model parameters which were constant in all the runs.
Table 5.1. Summary of the variable model parameters and runs. Five values of TSCF (0.0, 0.25, 0.50, 0.75, and 1.0) were used in each case. Width, WET m
WET/Ws Length, LET m
Qin (wells) m3/d/cell
Qin (total) m3/d
QET (max ET rate) m3/d/cell
ET area, Cells
QET (total) m3/d
100 1.0 1000 1.5 150 0.0005 20×200 50 150 1.5 1000 1.5 150 0.0005 30×200 75 200 2.0 1000 1.5 150 0.0005 40×200 100 250 2.5 1000 1.5 150 0.0005 50×200 125 300 3.0 1000 1.5 150 0.0005 60×200 150 100 1.0 500 1.50 150 0.001 20×100 50 150 1.5 500 1.50 150 0.001 30×100 75 200 2.0 500 1.50 150 0.001 40×100 100 250 2.5 500 1.50 150 0.001 50×100 125 300 3.0 500 1.50 150 0.001 60×100 150
117
Table 5.2. Constant Model Parameters.
Groundwater flow Parameter value Contaminant Transport Parameter
value
Horizontal Hydraulic Conductivity 187 m/d Substrate1 initial concentration 1.0 mg/L
Horizontal Anisotropy 1.0 Porosity 0.25 Root Extinction depth 4.0 m Longitudinal Dispersivity 10.0 m
Ground Surface Elevation 5.0 m Transverse/Longitudinal Dispersivity 0.2
Constant Head Boundary 5.25 m First order decay rate 0.001 1/d Model Thickness (one layer) 10.0 m Number of stress periods 20 Stress period length 182.5 d
The metrics by which different simulations will be compared are based on a quantitative reduction
in:
• Contaminant mass (plume) • Plume length (concentration based) • Mass flux:
o Downgradient of the source area o Other transects along the groundwater flow normal to plume centerline
118
5.3 Results and Discussions
5.3.1 Initial Test Case
In this section of the study, the new code is first compared with MT3DMS-SSM results. The model
is run using the previous parameters described in section 5.2 until he plume reaches a steady-state
under natural attenuation conditions only (no ET simulation). The concentration results from that
simulation are set to be the initial concentrations in the model where phytoremediation is simulated,
Figure 5.5.
NA Starting simulation Conditions NA End of simulation*
*ET Starting Simulation Conditions ET End of simulation
Figure 5.5. Initial conditions for the test models.
The phytoremediation system dimensions were 500m×300m (LET=0.5Lp, and WET is selected to be
three times the source width, Ws). The maximum ET rate was 0.001 m3/d/cell giving a total ET rate
equals to the total inflow rate = 150 m3/d. Figure (5.6.a) shows solute mass in the aquifer (or model
domain) and (5.6.b) shows the solute mass removed (sink term) for both MT3DMS-SSM and
SEAM3D-PUP with TSCF=1.0. The results indicated exact match. Figure 5.6.a. shows that the solute
mass in the aquifer reaches a constant value in case of natural attenuation conditions and after the
plume reached a steady state. The plot shows that the solute mass in the aquifer begins to oscillate up
and down according to the stress periods where ET is active and then reaches a dynamically steady
state, Figure 5.6.a.
119
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, gMT3DMS-SSM SEAM3D-PUP, TSCF=1.0 NA
0
20
40
60
80
100
120
140
160
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-ou
t(sin
ks),
g
MT3DMS-SSM SEAM3D-PUP NA
(a) (b)
Figure 5.6. Validation the results of SEAM3D-PUP by comparing the mass output of
MT3DMS-SSM versus PUP for a), solute mass in aquifer and b) solute mass removal for
LET=0.5Lp and WET=300m.
After validating the run results, additional model runs were performed for different values of TSCF
(0.0, 0.25, 0.50, 0.75) to check the results sensitivity of solute mass removal towards TSCF. The results
for this run are represented in the Figures (5.7.a and 5.7.b) which show the expected trend of increased
solute mass removal with higher values of TSCF. The increased mass-flux to the phytoremediation
zone occurs for all values of TSCF as the flowrate is increased no matter what the mass-removal is.
Also Figure 5.7.a shows that the larger the TSCF value, the longer will it take the plume to reach the
dynamic steady-state condition where the mass in the aquifer oscillates up and down a constant value.
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, g
SSM, W=300TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0NA
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, g
TSCF=0.0NA
(a) (b)
Figure 5.7. Solute mass in the model domain for a) Different values of TSCF, and (b) The
dynamically stable plume shows constant mass removal under NA conditions and oscillates
around this value for TSCF = 0.0. (WET/Ws=3.0, LET=0.5Lp).
120
In Figure 5.7b, even when TSCF=0.0 (theoretically, there is no mass removal from the system), the
pumping and cyclic effect of ET draws more flow to the source area and thus increase the mass flux
from the source area. Figure 5.8 displays the decline in groundwater head along the model longitudinal
centerline due to the phytoremediation effect (for the case where LET=0.5Lp, and WET=300). Increasing
the hydraulic gradient or q to the source area will result in increasing the total solute mass in the system
model, which explains the higher values of M (in case of TSCF=0.0) than those of natural attenuation.
The same increase in mass flux at source occurs for all values of TSCF (Figure 5.7a).
5
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
0 200 400 600 800 1000 1200 1400
Dist., m
Hyd
raul
ic H
ead,
m
No ETWith ET
Figure 5.8. Groundwater hydraulic head profile showing the effect of phytoremediation.
121
5.3.2 Effect of ET area (WET and LET) on contaminant mass removal
The model runs presented in Table 5.1 were performed. The main three parameters under
investigation were: WET: 5 different values, LET: 2 different values, and TSCF: 5 different values giving a
total of 5×2×5 = 50 model runs. The initial ET width was selected to contain the plume as it was a little
wider than the plume. It would not be good from the practical point of view to select a wider
phytoremediation system relative to the plume width. The wider the phytoremediation system, the
more mass-flux towards the ET zone, and this will result in a wider plume. This also will be discussed
in more details in Section 5.3.4. Figure 5.9 shows that the higher the TSCF value (as TSCF reaches 1.0
theoretically), the higher the solute mass removal from the aquifer. It is also interesting to refer to the
fact that the amount of water withdrawn is the same for all TSCF values, but the solute mass removed
form the model domain is different. Figure 5.9 also shows that the amount of solute mass removal for
the same value of TSCF is higher in case of LET=0.5Lp compared with LET=Lp despite the fact that the
total QET is the same (maximum ET rate is different). The full set of figures of this section is presented
in Appendix A.
122
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, g
ET, W=300TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
a) LET=Lp, QET = 0.0005 m3/d/m2 WET=300
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, gET, W=300TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
b) LET=0.5Lp, QET = 0.001 m3/d/m2 WET=300
Figure 5.9. Effect of ET width on solute mass removal for different values of TSCF: a)
LET=Lp and b) LET=0.5Lp.
123
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
a) LET=Lp, QET = 0.0005 m3/d/m2 TSCF=0.750
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
b) LET=0.5Lp, QET = 0.001 m3/d/m2 TSCF=0.750
Figure 5.10. Effect of TSCF on solute mass removal ET width values for: a) LET=Lp, and b)
LET=0.5Lp.
The results indicating the followings:
1- The width of the phytoremediation system, WET, is affecting the solute mass-removal. The
higher the width, the more mass is removed, Figure 5.10. However, the reduction in solute
mass is not noticeable for WET=150 to 300.
2- The density of trees closer to the contaminant source (The case where LET=0.5Lp) has higher
effect on solute mass removal. The higher the trees density closer to the contaminant source
(represented in maximum ET rate), the higher the solute mass removal, Figure 5.9.
3- The width of the ET area will have slight effect on the mass removal for different values of
TSCF. Figure 5.10 shows the mass removal curves for different ET widths, and the charts are
very close together except for WET=100.
124
5.3.3 Effect of ET Area on Plume Concentration
For the same model parameters described in section 5.2, the solute dissolved concentration values
are estimated using SEAM3D-PUP at selected observation points along the longitudinal model
centerline and downstream the source, (Figure 5.11). The observation points are 100 m apart and are
listed in Table 5.2. The solute dissolved concentrations at the observation points under natural
attenuation conditions only, are presented in Table 5.3.
500.0
1500.0
L
100.01100.0
ET
WET
Figure 5.11. Observation points for concentration profile.
Table 5.3. Observation cells (i, j, k)
i j k Distance from the source, m
NA concentration, mg/L
50 8 1 0.0 1.0 50 28 1 100.0 0.6308 50 48 1 200.0 0.3843 50 68 1 300.0 0.2309 50 88 1 400.0 0.1388 50 108 1 500.0 0.08392 50 128 1 600.0 0.05108 50 148 1 700.0 0.0313 50 168 1 800.0 0.0193 50 188 1 900.0 0.01197 50 208 1 1000.0 0.00747 50 228 1 1100.0 0.00468
125
The ET width is changed five times (WET/WS = 3.0, 2.5, 2.0, 1.5, and 1.0) where Ws is the width of
the source. The TSCF values were (1.0, 0.75, 0.50, 0.25, 0.0) and the concentration values are recorded
at the observation points at different stress periods. Figure 5.12 shows the effect of WET on the
reduction of plume concentration with time. The wider the photo zone, the lower the plume
concentration downstream the source. The case where LET=0.5Lp showed more concentration
reduction with time than the case of LET=Lp. The reduction in concentration is due to the fact that
maximum ET rate in the case of LET=0.5Lp double the maximum ET rate in case of LET=Lp. Although
the total QET is the same, but the withdrawal effect of the trees in the first case is higher because it is
closer to the source (cells of highest concentrations).
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
NAW=300W=250W=200W=150W=100
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
NAW=300W=250W=200W=150W=100
Distance from source = 500 Distance from source = 500
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Con
c., m
g/L
NAW=300W=250W=200W=150W=100
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Conc
., m
g/L
NAW=300W=250W=200W=150W=100
Distance from source = 1000 Distance from source = 1000 a) LET=Lp b) LET=0.5Lp
Figure 5.12. Concentration profiles at distances = 500, and 1000 downstream the source for
different values of WET for a) LET=Lp and b) LET=0.5Lp, where TSCF=1.0.
Figure 5.13 shows the results of concentration profile at different stress periods for TSCF=1.0, and
for a) LET=Lp and b) LET=0.5Lp. The results is showing that the reduction in plume length due to
phytoremediation is a slow process (takes years to significantly reduce the plume length) and also
shows that the reduction in plume concentration starts to take place at a certain distance downstream
126
the source (approximately equal to 300m in case of LET=Lp, and 200m for LET=0.5Lp). The reduction
in the plume length was larger in the case of LET=0.5Lp than in the case of LET=Lp. The full set of
charts for this case study is presented in Appendix A.
W=300, L(ET)=Lp
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+365t=+1825t=+3650
W=300, L(ET)=0.5Lp
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+365t=+1825t=+3650
a) b)
Figure 5.13. Concentration vs. distance at different observation points downstream the
source at the end of different stress periods for a) LET=Lp and b) LET=0.5Lp.
W/Ws=3.0, L(ET)=Lp
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
a) LET=Lp, QET = 0.0005 m3/d/m2
W/Ws=3.0, L(ET)=0.5Lp
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
b) LET=0.5Lp, QET = 0.001 m3/d/m2
Figure 5.14. Concentration profiles for different TSCF values used to calculate the plume
length at a concentration = 1% of the source concentration for a) LET=Lp and b) LET=0.5Lp.
127
The concentration profile charts (Figures 5.14) were used to find the distance at a concentration =
1% of the source concentration (i.e. 0.01 mg/L) and that distance is considered to be the plume length,
Lp* (The shrunk plume length due to phytoremediation effect).
The results are arranged in Table 5.4 and 5.5. The plume length under natural attenuation
conditions, Lp, and the reduced plume length due to the phytoremediation effect, Lp*, are used to
create deign charts to estimate the phytoremediation dimensions required to reach certain RAO. The
charts and design examples are presented in Section 5.4.
The concentration results indicating the followings:
1- The width of the phytoremediation system, WET, is affecting the solute concentration
downstream the source and thus the plume length. The higher the width of the
phytoremediation system, the shorter the plume length for both LET=0.5Lp and LET=Lp,
(Figure 5.12).
2- Figure 5.14 shows the sensitivity of plume concentration towards the range of TSCF values.
The trend was expected as the TSCF increases, the more the solute mass removal, and thus
decreasing the concentration. It is also noticeable that the concentration profiles for different
values of TSCF tend to gather in one line after a certain distance downstream the source in the
case of LET=0.5Lp. While in the case of LET=Lp, the concentration profiles for different values
of TSCF remain separate lines. This leads to the conclusion that TSCF will have minimal effect
of the concentration after a certain distance downstream the source, and the longer LET, the
longer that distance will be.
3- The density of trees closer to the contaminant source (The case where LET=0.5Lp) has higher
effect on the solute concentration reduction downstream the source. The higher the tree
density closer to the contaminant source (represented in maximum ET rate), the lower the
solute concentration downstream the source, (Figures 5.13 and 5.14). Comparing the length of
plume after phytoremediation, Lp*, and the plume length under natural attenuation conditions,
Lp, for different ET lengths is presented in Figures 5.15.
128
4- The reduction of the plume length relative to the NA plume length ( %*
p
pp
LLL −
) reached a
percentage of approximately 26% for the case where LET=0.5Lp or with maximum ET rate of
0.001 m3/d/m2) and a value of approximately 15% for the case where LET=Lp or with
maximum ET rate of 0.0005 m3/d/m2 after 10 years of applying the phytoremediation system,
(Figures 5.16 and 5.17).
The full set of figures of this section model runs are presented in Appendix A.
129
Table 5.4. Plume lengths at a concentration equals to 1% of the source concentration for ET length = 1000 m (approximately equals to the plume length).
ET width, WET
TSCF Qin (100 cells)
Max ET rate
ET area QET Uin Lp Lp* Lp*/Lp
UET =qET*LET, 1000*.001
UET/Uin
M m3/d/cell (m3/d)/m2 (cells) m3/d m2/d m m m2/d
300 1.00 1.5 0.0005 200×60 150 0.3 940 759 0.808 0.5 1.67
250 1.00 (150 m3/d) 0.0005 200×50 125 0.3 940 784 0.835 0.5 1.67
200 1.00 0.0005 200×40 100 0.3 940 805 0.856 0.5 1.67
150 1.00 0.0005 200×30 75 0.3 940 827 0.880 0.5 1.67
100 1.00 0.0005 200×20 50 0.3 940 850 0.904 0.5 1.67
300 0.75 1.5 0.0005 200×60 150 0.3 940 784 0.835 0.5 1.67
250 0.75 0.0005 200×50 125 0.3 940 810 0.862 0.5 1.67
200 0.75 0.0005 200×40 100 0.3 940 831 0.884 0.5 1.67
150 0.75 0.0005 200×30 75 0.3 940 857 0.912 0.5 1.67
100 0.75 0.0005 200×20 50 0.3 940 879 0.936 0.5 1.67
300 0.50 1.5 0.0005 200×60 150 0.3 940 809 0.861 0.5 1.67
250 0.50 0.0005 200×50 125 0.3 940 834 0.888 0.5 1.67
200 0.50 0.0005 200×40 100 0.3 940 859 0.913 0.5 1.67
150 0.50 0.0005 200×30 75 0.3 940 885 0.942 0.5 1.67
100 0.50 0.0005 200×20 50 0.3 940 905 0.963 0.5 1.67
300 0.25 1.5 0.0005 200×60 150 0.3 940 834 0.887 0.5 1.67
250 0.25 0.0005 200×50 125 0.3 940 862 0.917 0.5 1.67
200 0.25 0.0005 200×40 100 0.3 940 887 0.944 0.5 1.67
150 0.25 0.0005 200×30 75 0.3 940 912 0.970 0.5 1.67
100 0.25 0.0005 200×20 50 0.3 940 934 0.994 0.5 1.67
300 0.0 1.5 0.0005 200×60 150 0.3 940 858 0.912 0.5 1.67
250 0.0 0.0005 200×50 125 0.3 940 889 0.946 0.5 1.67
200 0.0 0.0005 200×40 100 0.3 940 912 0.970 0.5 1.67
150 0.0 0.0005 200×30 75 0.3 940 939 0.999 0.5 1.67
100 0.0 0.0005 200×20 50 0.3 940 960 1.021 0.5 1.67
130
Table 5.5. Plume lengths at a concentration equals to 1% of the source concentration for ET length = 500 m (approximately half the plume length).
ET width, WET
TSCF Qin (100 cells)
Max ET rate
ET area QET Uin Lp Lp* Lp*/Lp
UET =qET*LET, 500*.001
UET/Uin
m3/d/cell (m3/d)/m2 (cells) M3/d m2/d m m m2/d
300 1.00 1.5 0.001 100×60 150 0.3 940 628 0.668 0.5 1.67
250 1.00 (150 m3/d) 0.001 100×50 125 0.3 940 666 0.709 0.5 1.67
200 1.00 0.001 100×40 100 0.3 940 706 0.751 0.5 1.67
150 1.00 0.001 100×30 75 0.3 940 747 0.795 0.5 1.67
100 1.00 0.001 100×20 50 0.3 940 791 0.842 0.5 1.67
300 0.75 1.5 0.001 100×60 150 0.3 940 656 0.698 0.5 1.67
250 0.75 0.001 100×50 125 0.3 940 692 0.736 0.5 1.67
200 0.75 0.001 100×40 100 0.3 940 737 0.784 0.5 1.67
150 0.75 0.001 100×30 75 0.3 940 778 0.828 0.5 1.67
100 0.75 0.001 100×20 50 0.3 940 821 0.873 0.5 1.67
300 0.5 1.5 0.001 100×60 150 0.3 940 676 0.719 0.5 1.67
250 0.5 0.001 100×50 125 0.3 940 718 0.764 0.5 1.67
200 0.5 0.001 100×40 100 0.3 940 762 0.811 0.5 1.67
150 0.5 0.001 100×30 75 0.3 940 802 0.853 0.5 1.67
100 0.5 0.001 100×20 50 0.3 940 845 0.899 0.5 1.67
300 0.25 1.5 0.001 100×60 150 0.3 940 695 0.740 0.5 1.67
250 0.25 0.001 100×50 125 0.3 940 737 0.784 0.5 1.67
200 0.25 0.001 100×40 100 0.3 940 786 0.836 0.5 1.67
150 0.25 0.001 100×30 75 0.3 940 830 0.883 0.5 1.67
100 0.25 0.001 100×20 50 0.3 940 872 0.928 0.5 1.67
300 0.0 1.5 0.001 100×60 150 0.3 940 713 0.759 0.5 1.67
250 0.0 0.001 100×50 125 0.3 940 758 0.806 0.5 1.67
200 0.0 0.001 100×40 100 0.3 940 806 0.857 0.5 1.67
150 0.0 0.001 100×30 75 0.3 940 849 0.903 0.5 1.67
100 0.0 0.001 100×20 50 0.3 940 893 0.950 0.5 1.67
131
TSCF=1.0
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
W/Ws
Lp*/L
p L(ET)=0.5Lp
L(ET)=Lp
W/Ws=3.0
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
0 0.25 0.5 0.75 1
TSCF
Lp*/L
p L(ET)=0.5LpL(ET)=Lp
Figure 5.15. Comparison of the plume length under ET, (Lp*), to the plume length under
natural attenuation only, (Lp), for different ET dimensions (W/Ws) and TSCF values.
132
W=300, L(ET)=0.5Lp, t=+365, TSCF=1.0
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+365NA (+365)
W=300, L(ET)=Lp, t=+365, TSCF=1.0
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L
t=+365NA (+365)
(a) Time = +365 after phytoremediation starts
Figure 5.16. Concentration profiles at different times after the phytoremediation system
starts for two different LET.
133
W=300, L(ET)=0.5Lp,t=+1825, TSCF=1.0
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+1825NA
W=300, L(ET)=Lp, t=+1825, TSCF=1.0
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+1825NA
(b) Time = +1825 (5 years) after phytoremediation starts
Figure 5.16. Concentration profiles, Continued.
134
W=300, L(ET)=0.5Lp,t=3650, TSCF=1.0
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+3650NA
L(ET)=Lp, W=300, t=3650, TSCF=1.0
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+3650NA
(c) Time = +3650 (10 years) after phytoremediation starts
Figure 5.16. Concentration profiles, continued.
135
0.0
5.0
10.0
15.0
20.0
25.0
30.0
0.00 730.00 1460.00 2190.00 2920.00 3650.00
+Time, d
% r
educ
tion
in p
lum
e le
ngth
L(ET)=0.5LpL(ET)=Lp
Figure 5.17. Reduction in plume length due to phytoremediation.
136
5.3.3.1 Radioactive decay or biodegradation
The groundwater contaminant chemical reactions include first-order irreversible rate reaction (such
as radioactive decay), reversible equilibrium-controlled sorption with linear, Freundlich, or Langmuir
isotherms, and reversible equilibrium-controlled ion exchange for divalent ions, (Zheng and Wang,
1998).
The first-order irreversible rate reaction term included in the governing equation, ( )CC bρλθλ 21 + ,
represents the mass loss of both the dissolved phase (C) and the sorbed phase ( )C . The rate constant
is usually given in terms of the half-life, ⎟⎟⎟
⎠
⎞
⎜⎜⎜
⎝
⎛=
21
2lnt
λ , (Fetter 1999).
Where t1/2 is the half-life of radioactive or biodegradable materials (i.e., the time required for the
concentration to decrease to one-half of the original value).
For radioactive decay, the reaction generally occurs at the same rate in both phases. For
biodegradation, however, it has been observed that certain reactions occur only in the dissolved phase.
That is why two different rate constants may be needed. It should be noted that various biodegradation
processes in the subsurface are usually more complex than that described by the first-order irreversible
rate reaction, (Zheng and Wang, 1998).
The radioactive decay/degradation process is assumed to follow first-order kinetics, which means
that the rate of loss of mass at any given time is directly proportional to the mass present at that time.
The contaminant concentration at a distance x relative to the source concentration, is given in the
equation: ( )⎥⎥
⎦
⎤
⎢⎢
⎣
⎡
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛ +−= x
DDvv
CxCL
Lxx
24
exp2
0λ
, (Bedient et al., 1994).
It is estimated that enhanced biodegradation will occur in the plantation area. Figure 5.18 represents
the contaminant concentration profile for different values of decay rate, λ, which indicates that the
higher the value of λ, the lower the plume concentration and thus the shorter the plume length.
Increasing the decay rate from 0.001 to 0.002 d-1 (to resemble the rhizosphere effect) resulted in
137
decreasing the plume length from 625m to 400m approximately which represents about 36%
reduction.
W/Ws=3.0, L(ET)=0.5Lp, TSCF=1.0
0.00001
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NALamda=0.001Lamda=0.002
NA
λ=0.001 d-1
λ=0.002 d-1
Figure 5.18. Effect of decay rate due to phytoremediation on the dissolved concentration.
138
5.3.4 Effect of ET area and TSCF on mass-flux
The mass-flux of the contaminants is equal to the average concentration (mg/L) times the average
flowrate (L/d). The flowrate is calculated for each cell along the model cross-section using the equation
( )L
hhKAQ 12 −−= , Where Q is the volumetric flow (L3T-1); K is the hydraulic conductivity of the
material in the direction of flow (LT-1); A is the cross-sectional area perpendicular to the flow (L2); h1-h2
is the head difference across the prism parallel to flow (L); and L is the length of the prism parallel to
the flow path (L).
The flowrate values of each cell can be found in the binary MODFLOW output file with extension
*.CCF (the cell to cell flow file). The value of right-face flow (flow leaving the cell) is found for each
cell, and multiplied by the concentration at the same cell to find the mass-flux at that particular cell.
The average mass-flux will equal to the summation of the mass-flux of all the cells at a certain cross-
section, , (Figure 5.19). ∑ ×=n
ii qCm&
q (in)x
Δx
q (out)x
ΔyΔ
z
x x+Δx
hh
12
Cn
Figure 5.19. Calculating of mass-flux for the flow model of SEAM3D-PUP.
C1 C2
C3
qn
∑ ×=n
ii qCm&
q1 q2
q3
139
The model parameters are the same as listed in table 5.1. Figure 5.20.a shows the flowrate leaving
the right cell face (on the left y axis), the transverse concentration profile at X=500 m downstream the
source, and the contaminant mass-flux (on the right y-axis). The contaminant mass-flux is maximum at
the model centerline, and decreases at both ends to the left and right of the flow direction. Figure
5.20.b shows the sensitivity of mass-flux results towards TSCF for a model width equal to 300.0 m
(WET/Ws= 3.0) and LET=Lp. The higher the TSCF, the lower the solute mass-flux.
The mass-flux was reduced even when TSCF=0.0 due to the groundwater withdrawal by the trees.
Figure 5.21.a shows a reduction in mass-flux due to the phytoremediation system relative to the mass-
flux of natural attenuation conditions only, in a magnitude of 97% in case of TSCF=1.0. Comparing
the mass-flux distribution across the model width (normal to the flow direction) at a distance = 500 m
for different values of ET widths, shows that the mass-flux reduction in case of LET=0.5Lp is greater
than that of LET=Lp. The two dashed vertical lines in the charts represent the left and right boundaries
of the ET area, Figures 5.20.a and 5.21.a. The phytoremediation system was effective to the extent that
it reversed the mass-flux for the case where WET=300 shown in 5.20b.
All the rest of the run figures for different values of ET width and length, and different TSCF
values are in Appendix A.
140
X=500, L(ET)=0.5Lp
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 10 20 30 40 50 60 70 80 90 100
cell # across the model width
Flow
, m3/
d
0
0.007
0.014
0.021
0.028
0.035
0.042
0.049
Con
c., m
g/L
& M
ass-
flux,
g/d
FlowConc.Mass-flux
(a)
W=300, X=500, L(ET)=Lp
-10
10
30
50
70
90
110
130
150
0 10 20 30 40 50 60 70 80 90 100
Cell # across the model width
Mas
s-flu
x, m
g/d TSCF=1.0
TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0NA
(b)
Figure 5.20. Distribution of right-face cell flow (out-flow), aqueous concentration and mass-
flux at a cross-section 500 m DS the source (WET/WS = 2.0).
141
W=250, X=500, L(ET)=0.5Lp
0
2
4
6
8
10
12
14
0 10 20 30 40 50 60 70 80 90 100
Cell # across the model width
Mas
s-flu
x, m
g/d
0
20
40
60
80
100
120
140
Mas
s-flu
x (N
A), m
g/d TSCF=1.0
TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0NA
(a)
W=300, X=500, L(ET)=0.5Lp
-10
-8
-6
-4
-2
0
2
4
0 20 40 60 80 100
Cell # across the model width
Mas
s-flu
x, m
g/d
-10
10
30
50
70
90
110
130
Mas
s-flu
x (N
A),
mg/
d TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0NA
Figure 5.21. Mass-flux distribution at a cross-section 500 m DS the source for different TSCF
values for a) WET/WS =2.50), and b) WET/WS =3.0.
142
The average mass-flux values were then calculated at different model cross-sections 100m apart.
The cell concentrations, Ci, and the cell flowrate, qi are estimated at each cell along the model cross-
section. The summation gives the average mass-flux at a particular cross-section. ∑ ×=n
ii qCm&
The average mass-flux results represented in Figure 5.22 indicated the efficiency of using a
phytoremediation system to reduce the contaminant mass-flux. The system with dense trees (and
higher QET of LET =0.5Lp) was more efficient in mass-flux reduction at all downstream sections and
even reversed the mass-flux direction in the case of WET/Ws=3.0, Figure 5.21.b. The negative
numerical values of mass-flux are not clear in Figure 5.22.b, but the full results of the average mass-flux
for all the model runs in this section can be found in Table A.1, Appendix A.
-5
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Thou
sand
s
Dist., m
Av.,
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
a) LET=Lp, QET = 0.0005 m3/d/m2 WET=300
-5
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100
Thou
sand
s
Dist., m
Av.,
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
b) LET=0.5Lp, QET = 0.001 m3/d/m2 WET=300
Figure 5.22. Average Mass-flux results at different cross-sections downstream of the source
for a) LET=Lp and b) LET=0.5Lp for different values of TSCF, and WET=300.
143
The average mass-flux results indicated the following: 1) The highest reduction in mass-flux
occurred for the highest value of WET and TSCF, Figure 5.23; 2) The reduction in mass-flux is
proportionally increasing with the increase of ET width in the case where LET = Lp, and changes
abruptly after the ET width is larger then the source width in the case where LET=0.5Lp, (Figure 5.24).
-5
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Thou
sand
s
Dist., m
Av.
, Mas
s-flu
x, m
g/d
NAL(ET)=LpL(ET)=0.5Lp
Figure 5.23. Average contaminant mass-flux at different cross-sections downstream the
source for LET=Lp and LET=0.5Lp, (WET=300, and TSCF=1.0).
X=500, TSCF=1.0, L(ET)=Lp
0
500
1000
1500
2000
2500
3000
3500
0 0.5 1 1.5 2 2.5 3
W(ET)/Ws
Av. M
ass-
flux,
mg/
d
0
500
1000
1500
2000
2500
3000
3500
Av. d
iff in
mas
s-flu
x, m
g/d
Av. MFAv. Diff. MF
X=500, TSCF=1.0, L(ET)=0.5Lp
-100
400
900
1400
1900
2400
2900
3400
0 0.5 1 1.5 2 2.5 3
W(ET)/WS
Av. M
ass-
flux,
mg/
d
-100
400
900
1400
1900
2400
2900
3400
Av.
diff
in m
ass-
flux,
mg/
d
Av. MFAv. Diff. MF
Figure 5.24. Average mass-flux reduction vs. (W/Ws) for different values of TSCF and LET.
144
5.4 Effect of Groundwater Flux and ET flux rates
The second set of runs will investigate the effect of the model system (or the unconfined aquifer) in-
flux represented in the well discharge, Qin relative to the out-flux represented in QET. The in-flux is a
function of the saturated thickness, h, that is ( ) ⎟⎟⎟
⎠
⎞
⎜⎜⎜
⎝
⎛=
×==×=
md
m
BQ
hBQh
AQhqU ininin
in
3
, where h is
the saturated thickness and B is the model width. The in-flow to the aquifer is kept constant by using
injection wells at the left boundary. The total flow-in will equal to the number of wells multiplied by
the well flow. Three values for well flow are assumed (2.0, 1.5, and 1.05 m3/d/cell) giving three
different values of aquifer flux (0.4, 0.3, and 0.21 m3/d/m). On the other hand, the out-flux of the
aquifer resulted from the phytoremediation system , where is the maximum ET
rate which is kept constant and equal to 0.0005 m3/d/m2 and LET is the phytoremediation system
length, m. The out-flux values are controlled by changing LET values. LET is selected according to the
plume length under natural attenuation conditions, (Figure 5.25a).
ETETET LqU ×= max maxETq
After the model is run under the previous conditions, and the plume is already characterized
according to geological and hydro-geological parameters reaching steady-state stability, the length of
ET is changed four times in proportion to the recorded stable plume length to be equal to (Lp, 0.75Lp,
0.5Lp, and 0.25Lp), (Figure 5.25.b). Furthermore, the phytoremediation area is placed at different
locations in the model relative to the contaminant source and the plume toe. If the phytoremediation
area starts at the plume toe going towards the source, a model parameter defining the distance from the
source to the phyto zone, XET is introduced, (Figure 5.25.c). When the phytoremediation zone starts at
the contaminant source, XET will equal to zero. The total number of model runs is displayed in Tables
5.6 and 5.7.
145
LET
WET
Qin
QET
Constant-head Cells
(a)
L(ET)1L(ET)2
L(ET)3
L(ET)4
(b)
Source
In-Flow
ET
XET (c)
Figure 5.25. Conceptual model for the study case 5-4.
146
Table 5.6. Phytoremediation area starts at the source (XET=0.0). NA length, Lp=1237, ET Max. Length=1240, TSCF=1.0, WET=300, ET max rate=0.0005 m3/d/cell ET length, LET
Area m2
Qin m3/d/cell
Qin m3/d
QET m3/d
Uin m2/d
Lp m
Lp* m
Lp*/Lp
UET m2/d
UET/Uin
310 93000 2.0 200 46.5 0.4 1237 1125.4 0.910 0.16 0.388 620 186000 2.0 200 93 0.4 1237 1044.4 0.844 0.31 0.775 930 279000 2.0 200 139.5 0.4 1237 1009.2 0.816 0.47 1.163 1240 372000 2.0 200 186 0.4 1237 1011.4 0.818 0.62 1.550
NA length = 976, ET max length = 980, TSCF=1.0, WET=300, ET max rate = 0.0005 ET length, LET
Area m2
Q inm3/d/cell
Qin m3/d
QET m3/d
U inm2/d
Lp m
Lp* m
Lp*/Lp
U T Em2/d
UET/Uin
245 73500 1.5 150 36.75 0.3 976 886.2 0.908 0.12 0.408 490 147000 1.5 150 73.5 0.3 976 820.7 0.841 0.25 0.817 735 220500 1.5 150 110.25 0.3 976 793.4 0.813 0.37 1.225 980 294000 1.5 150 147 0.3 976 796.5 0.816 0.49 1.633
NA length = 703.5, ET max length = 700, TSCF=1.0, WET=300, ET max rate = 0.0005 ET length, LET
Area m2
Qin m3/d/cell
Qin m3/d
QET m3/d
Uin m2/d
Lp m
Lp* m
Lp*/Lp
UET m2/d
UET/Uin
175 52500 1.05 105 26.25 0.21 731.5 664.6 0.909 0.09 0.417 350 105000 1.05 105 52.5 0.21 731.5 614.3 0.840 0.18 0.833 525 157500 1.05 105 78.75 0.21 731.5 592.0 0.809 0.26 1.250 700 210000 1.05 105 105 0.21 731.5 594.8 0.813 0.35 1.667 Table 5.7. Phytoremediation area starts at the plume toe (XET is variable) NA length = 976, ET max length = 980, TSCF=1.0, WET=300, ET max rate = 0.0005 ET length, LET
XET Area m2
Q inm3/d/cell
Qin m3/d
QE T m3/d
U inm2/d
Lp m
Lp* m
Lp*/Lp
U T Em2/d
UET/Uin
245 0.75Lp 73500 1.5 150 36.75 0.3 976 968.7 0.992 0.12 0.408 490 0.5Lp 147000 1.5 150 73.5 0.3 976 933.4 0.956 0.25 0.817 735 0.25Lp 220500 1.5 150 110.25 0.3 976 873.8 0.895 0.37 1.225 980 0.0 294000 1.5 150 147 0.3 976 796.5 0.816 0.49 1.633 NA length = 1237, ET max length = 1240, TSCF=1.0, WET=300, ET max rate = 0.0005 ET length, LET
XET Area m2
Qin m3/d/cell
Qin m3/d
QET m3/d
Uin m2/d
Lp m
Lp* m
Lp*/Lp
UET m2/d
UET/Uin
310 0.75Lp 93000 2.0 200 46.5 0.4 1237 1225.6 0.991 0.16 0.388 620 0.5Lp 186000 2.0 200 93 0.4 1237 1181.0 0.955 0.31 0.775 930 0.25Lp 279000 2.0 200 139.5 0.4 1237 1107.3 0.895 0.47 1.163 1240 0.0 372000 2.0 200 186 0.4 1237 1011.4 0.818 0.62 1.550 NA length =703.5, ET max length = 700, TSCF=1.0, WET=300, ET max rate = 0.0005 ET length, LET
XET Area m2
Qin m3/d/cell
Qin m3/d
QET m3/d
Uin m2/d
Lp m
Lp* m
Lp*/Lp
UET m2/d
UET/Uin
175 0.75Lp 52500 1.05 105 26.25 0.21 731.5 723.7 0.989 0.09 0.417 350 0.5Lp 105000 1.05 105 52.5 0.21 731.5 696.0 0.951 0.18 0.833 525 0.25Lp 157500 1.05 105 78.75 0.21 731.5 651.6 0.891 0.26 1.250 700 0.0 210000 1.05 105 105 0.21 731.5 594.8 0.813 0.35 1.667
147
5.4.1 Effect of Aquifer In-Flux/ Out-Flux on Mass Removal
The solute mass removal (represented in total solute mass in the aquifer) is presented in the figures
5.26 to 5.29. The starting point of simulation is when phytoremediation is applied after the plume has
reached a steady-state. The total simulation time after phytoremediation effect is active is ten years or
twenty stress periods. Depending on the value of in-flux, Qin, the length of the steady-state plume is
determined. Results showed that the steady-state plume lengths are 1237m, 976m, and 731.5m for
Qin= 200, 150, 105 m3/d respectively, (Table 5.4). The longest ET lengths (LET=Lp) are then selected
to be 1240m, 980m, and 700m respectively.
Figure 5.26 shows the solute mass in the aquifer (or model domain) for different aquifer in-flux
rates (200, 150, and 105 m3/d) and four different ET lengths (LET=Lp, 0.75Lp, 0.5Lp, and 0.25Lp). The
different ET lengths produce different out-flux, UET. The values of UET are presented in table 5.4 and
5.5. For each of the ET lengths, two different locations for the phyto area are selected. The first
placement is at the source (the left edge of the phyto area coincide with the source left edge), and at the
plume toe (The right edge of the phyto area is touching or slightly to the right of the plume toe). The
two previous phyto locations will be referred to as: (at source, and at the plume toe), respectively. For
all the runs in this section, TSCF value was assumed to be 1.0.
Also, Figures 5.26 and 5.27 are showing that the placement of the ET area away from the
contaminant source has very low effect on the solute mass removal even though the quantity of
groundwater transpired is the same. For example, comparing the location of the ET areas in Figure
5.28 is showing that placing a phyto system of LET = 0.5Lp starting the contamination source gave
much better results for solute mass removal than placing the same ET area at the plume toe.
148
LET=0.25 Lp
LET= 0.50Lp
LET= 0.75Lp
LET= Lp
Q=200, ET at the source
2022242628303234363840424446
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s rm
oval
, g L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.00
(a) (b) Q=150, ET at the source
2022242628303234363840424446
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s rm
oval
, g L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.00
Q=105, ET at the source
2022242628303234363840424446
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s rm
oval
, g L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.00
(c) (d)
Figure 5.26. Solute mass in the aquifer (or model domain) for different aquifer in-flux and
ET lengths (different out-flux) where the ET length starts at the source, TSCF=1.0.
149
LET=0.25 Lp
LET= 0.50Lp
LET= 0.75Lp
LET= Lp
Q=200, ET at the plume toe
20
25
30
35
40
45
50
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s rm
oval
, g L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.00
(a) (b) Q=150, ET at the plume toe
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s rm
oval
, g L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.00
Q=105, ET at the plume toe
20
20.5
21
21.5
22
22.5
23
23.5
24
24.5
25
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s rm
oval
, g L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.00
(c) (d)
Figure 5.27. Solute mass in the aquifer (or model domain) for different aquifer in-flux and
ET lengths (different out-flux) where the ET length starts at the plume toe.
L(ET)/Lp=0.50
28
29
30
31
32
33
34
35
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s re
mov
al, g
ET at sourceET at plume toe
L(ET)/Lp = 0.75
25
26
27
28
29
30
31
32
33
34
35
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s re
mov
al, g
ET at sourceET at plume toe
Figure 5.28. Comparison of solute mass in aquifer for different ET placement.
150
There is a trend of increased solute mass removal with the increase of the ET length relative to the
plume length, LET/Lp. The reduction of solute mass is acute for values of LET/Lp < 1.0 and tends to
reach a stable value with LET/Lp > 1.0 which leads to the conclusion of the closer the ET area to the
contaminant source, the more efficient the system for mass removal, figure 5.29.a and b.
Figures 5.29.c and 5.29.d show that the change in aquifer in-flux, Qin, has minimal effect on the
solute mass reduction when the phyto zone is at the plume toe.
ET at the source
10
11
12
13
14
15
16
17
18
0 0.2 0.4 0.6 0.8 1 1.2
L(ET)/Lp
% re
duct
ion
in s
olut
e m
ass
Qin=200Qin=150Qin=105
ET at the source
10
11
12
13
14
15
16
17
18
0.000 0.500 1.000 1.500 2.000
U(ET)/U(in)
% re
duct
ion
in s
olut
e m
ass
Qin=200Qin=150Qin=105
(a) (b) ET at the plume toe
0
2
4
6
8
10
12
14
16
18
0 0.2 0.4 0.6 0.8 1 1.2
L(ET)/Lp
% re
duct
ion
in s
olut
e m
ass
Qin=200Qin=150Qin=105
ET at the plume toe
0
2
4
6
8
10
12
14
16
18
0.000 0.500 1.000 1.500 2.000
U(ET)/U(in)
% r
educ
tion
in s
olut
e m
ass
Qin=200Qin=150Qin=105
(c) (d)
Figure 5.29. Effect of out-flux, UET relative to in-flux, Uin on the solute mass removal.
151
5.4.2 Effect of aquifer in-flux/ out-flux on plume concentration
The next parameter in the design outcomes is the reduction in plume concentration due to using of
a phytoremediation system. The concentration profiles for different ET lengths and locations is shown
in Figures 5.30 for the value of in-flux, Qin=200. The rest of the charts for Qin=150, 105 m3/d are
shown in Appendix A.
The in-flow discharge to the model domain is controlled by changing the well flowrate at the model
left boundary. The results indicated that the higher the in-flux, the longer the plume length to reach the
dynamic stability. The shorter ET lengths (for the cases of LET=0.25 Lp) has little effect on the plume
concentration specially if its location is at the plume toe, (Figure 5.30). The concentration profiles for
all the ET lengths and locations are lower than the natural attenuation-only concentration.
Comparing the effect of ET location on the plume downstream concentration indicated that the
best location for a phytoremediation system is closer as possible to the contaminant source, (Figure
5.31). TSCF has the expected effect on the plume length that is the higher the TSCF value, the more
solute mass is uptaken, and the lower the plume concentration.
152
Q=200
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200 1400
Con
c., m
g/L
Dist., m
L(ET)=0.25 LpL(ET)=0.50 LpL(ET)=0.75 LpL(ET) = LpNA
(a) ET starts at the source
Q=200
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200 1400
Dist., m
Con
c., m
g/L
L(ET) = 0.25 LpL(ET) = 0.50 LpL(ET) = 0.75 LpL(ET) = LpNA
(b) ET starts at the plume toe
Figure 5.30. Concentration profiles for aquifer in-flux (Qin=2.0 m3/d/cell) and different ET
lengths and locations.
153
Q=150
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Con
c., m
g/L
L(ET)=0.25 Lpat right edge
Q=150
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Con
c., m
g/L
L(ET)=0.50 Lpat right edge
Q=150
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Con
c., m
g/L
L(ET)=0.75 Lpat right edge
Figure 5.31. Comparison for concentration profiles for different ET locations.
154
5.4.3 Effect of Aquifer In-Flux/Out-Flux on Average Solute Mass-Flux
The third parameter to investigate in the design outcomes is the mass-flux. The mass-flux results are
estimated for different inflow rates (200, 150, and 105 m3/d), for different ET lengths, and locations.
The ET lengths are changed four times with respect to the stable plume length: 0.25, 0.50, 0.75, and 1.0
of Lp. The steady-state plume length is different for each inflow rate, and LET are selected accordingly,
(Table 5.6 and 5.7). The first set of figures, Figure 5.32 to 5.35 displays the average solute mass-flux for
different LET lengths and locations, the reduction in mass-flux due to the effect of ET, and comparison
between mass-flux results in the case of the ET at the source and at the plume toe.
Figure 5.32 shows the average solute mass-flux at different cross-sections downstream the source.
The Qin for this chart equals to 200 m3/d, and a TSCF = 1.0. The mass-flux curves tend to be one
line at a distance closer to the source, and then each line separates a different downstream distance
according to LET. The lowest mass-flux values were for the case of LET = Lp (UET/Uin=1.55). It is clear
that the higher UET/Uin, the lower the mass-flux at the same cross-section.
Figure 5.33 shows the reduction and percentage reduction in mass-flux relative to the mass-flux
under natural attenuation conditions due to the use of a phyto system. The highest reduction in mass-
flux occurred at the plume toe for LET=Lp.
A comparison between mass-flux results for LET/Lp = 0.5 and 0.75 for a phytoremediation system
at the source and at the plume toes shows that the mass-flux is lower in the first case, (Figure 5.34.)
Figure 5.35 displays the mass-flux sensitivity to TSCF for different ranges of ET dimensions and
locations. The TSCF effect on mass-flux is almost insignificant for smaller values of UET/Uin and when
the ET area is at the plume toe.
155
Qin=200, ET at the source
1.0
10.0
100.0
1000.0
10000.0
100000.0
0 200 400 600 800 1000 1200 1400
Dist., m
Mas
s-flu
x, m
g/d L(ET)/Lp=0.25
L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0NA
Qin=200, ET at the plume toe
1.0
10.0
100.0
1000.0
10000.0
100000.0
0 200 400 600 800 1000 1200 1400
Dist., m
Mas
s-flu
x, m
g/d L(ET)/Lp=0.25
L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0NA
Figure 5.32. Average solute mass-flux for different LET lengths and locations, Qin=200 m3/d.
Reduction in mass flux
Qin=200, ET at the left edge
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Red
uctio
n in
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Reduction in mass fluxQin=200, ET at the right edge
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Red
uctio
n in
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
% Reduction in mass flux
Qin=200, ET at the left edge
-5
15
35
55
75
95
0 200 400 600 800 1000 1200 1400
Dist., m
% re
duct
ion
inM
ass-
flux
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
% Reduction in mass fluxQin=200, ET at the right edge
-5
15
35
55
75
95
0 200 400 600 800 1000 1200 1400
Dist., m
% r
educ
tion
in M
ass-
flux
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Figure 5.33. Average reduction in solute mass-flux (with respect to the NA conditions) for
different LET lengths and locations, Qin=200 m3/d.
156
Qin=200, L(ET)/Lp=0.50
1.0
10.0
100.0
1000.0
10000.0
100000.0
0 200 400 600 800 1000 1200 1400
Dist., m
Mas
s-flu
x, m
g/d
at sourceat plume toe
Qin=200, L(ET)/Lp=0.75
1.0
10.0
100.0
1000.0
10000.0
100000.0
0 200 400 600 800 1000 1200 1400
Dist., m
Mas
s-flu
x, m
g/d
at sourceat plume toe
L(ET)=0.5Lp
-1000.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Dist., m
Mas
s-flu
x (d
iffer
ence
), m
g/d
L(ET)=0.75Lp
-1000.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Dist., m
Mas
s-flu
x (d
iffer
ence
), m
g/d
Figure 5.34. Comparison between mass-flux results for different phytoremediation system
dimensions and locations.
157
Qin = 150, LET/Lp = 0.50.
0
5
10
15
20
25
30
35
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/da
y TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0NA
10
100
1000
10000
100000
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/da
y TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0NA
Semi-log scale Qin = 150, LET/Lp = 0.50.
0
5
10
15
20
25
30
35
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/da
y TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0NA
10
100
1000
10000
100000
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/da
y TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0NA
Semi-log scale
Figure 5.35. Effect of TSCF on the reduction of solute mass-flux (compared to the NA
conditions) for left and right locations of ET.
158
5.5 Effect of Dividing the ET Area into Two Halves
Sometimes the contaminant site may have constraints on selecting the location of a
phytoremediation system. It may not be possible to place the phyto area closer to the contaminant
source. The set of runs in this section is examining the effect of dividing the total ET area into two
halves and thus reporting the effect on the design metric (concentration, solute mass removal, and
mass-flux). The ET area is divided into two haves and the segments are arranged as shown in the
following figure:
1- First and third quarters (1 & 3) 3- One segment at the source
2- Second and fourth quarters (2 & 4) 4- One segment at the plume toe
The best performance for the phyto system, in terms of solute-mass removal and reducing the
plume concentration downstream the source, was the location #3 (one segment at the source), (Figure
5.36). Still if this selection is not available, the second best performance was the location #1 then #2
and location #4 comes in the last order. Figure 5.38 shows the order of the phyto system performance
in terms of solute mass removal to be as follows: Location #3, #1, #2, and then #4.
Figure 5.37 shows the results of solute mass-flux for the four arrangements listed above. The order
of best performance is the same as in mass removal and solute concentration.
159
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Con
c., m
g/L Right edge
2 & 41 & 3Left edge
28000
29000
30000
31000
32000
33000
34000
35000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Solu
te M
ass,
g Right Edge2 & 41 & 3Left Edge
(a) Plume concentration at the end of simulation
(b) Solute mass removal
Figure 5.36. Effect of splitting the ET area into two halves on solute concentration and mass
removal.
0
5000
10000
15000
20000
25000
30000
35000
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
Right Edge2 & 41 & 3Left Edge
1
10
100
1000
10000
100000
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
Right Edge2 & 41 & 3Left Edge
Figure 5.37. Effect of splitting the ET area into two halves on solute mass-flux.
160
0
2
4
6
8
10
12
14
16
18
Left Edge Right Edge 1&3 2&4
Redu
ctio
n of
sol
ute
mas
s %
Figure 5.38. % Reduction in solute mass for different ET arrangements.
161
5.6 Effect of Removing the Source
In this set of runs, the source will be removed by assigning zero concentration in the source area
and removing the constant concentration ID in the source cells. Different positions for the ET area
will be applied depending on the distance from the source. The object of this section is to evaluate the
usefulness of a phyto system even after the contaminant source is removed. The previous set of ET
area locations was used in the simulation of this section.
Figure 5.39 shows the concentration profiles along the model centerline at different time steps after
the contaminant source is removed. The profiles indicated that the use of a phyto system reduced the
solute concentration for the distance closer to the source to approximately 500m downstream the
source, and then follow the same trend of natural attenuation conditions.
The location of the phyto system has noticeable effect on the concentration profile after the source
is removed. Comparing the concentration profile for using the phyto system and under natural
attenuation only shows that the reduction in concentration zone changes with time. Figure 5.40 shows
that the reduction in concentration zone is at approximately 450m downstream the source at
t=+1825d, but started at approximately 820m downstream the source at time =+3650d. Figure 5.41
shows the same trend for LET=Lp. The reduction in solute concentration (after the source is removed)
for different LET lengths and locations is shown in Figure 5.42 which indicates that there was some
distance where the concentration due to the use of a phyto system will be less than the natural
attenuation concentration, but not for the whole distance downstream the source.
In terms of solute mass removal, applying a phyto system after the source is removed has good
effect on removing the contaminant in less time, (Figure 5.43). The solute mass reduction due to the
phytoremediation system where the contaminant source is removed for different ET dimensions (LET=
Lp, 0.5Lp at the source, and 0.5Lp at the plume toe) is presented in Figure 5.44.
162
Contaminant source removed
Removed source Vs. NA
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAt=+182.5t=+547.5t=+912.5t=+1277.5t=+1642.5
L(ET)=0.5Lp at the Right edge
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 200 400 600 800 1000 1200
Dist., m
Con
c., m
g/L
NAt=+182.5t=+547.5t=+912.5t=+1277.5t=+1642.5
L(ET)=Lp
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 200 400 600 800 1000 1200
Dist, m
Con
c., m
g/L
NAt=+182.5t=+547.5t=+912.5t=+1277.5t=+1642.5
L(ET)=0.5Lp at the left edge
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAt=+182.5t=+547.5t=+912.5t=+1277.5t=+1642.5
Figure 5.39. Concentration profiles at different time steps after the contaminant source is
removed.
163
L(ET)=0.5Lp (LEFT), t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
L(ET)=0.5Lp (LEFT), t=+3650
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAET
L(ET)=0.5Lp (RIGHT), t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
L(ET)=0.5Lp (RIGHT), t=+3650
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
Figure 5.40. Solute concentration profiles, source removed for LET=0.5Lp at left and right
sides of the plume footprint.
164
L(ET)=Lp, t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
L(ET)=Lp, t=+1825
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAET
t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NA L(ET)=0.5Lp (Right) L(ET)=0.5Lp (Left) L(ET)=Lp
t=+3650
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NA L(ET)=0.5Lp(Right) L(ET)=0.5Lp(Left) L(ET)=Lp
Figure 5.41. Solute concentration profiles, source removed for LET=Lp, and comparison of the
LET location effect on concentration.
165
t=+1825
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Red
uctio
n in
con
cent
ratio
n,
mg/
L
L(ET)=0.5Lp(LEFT) L(ET)=0.5Lp(RIGHT) L(ET)=Lp
t=+3650
-0.006
-0.005
-0.004
-0.003
-0.002
-0.001
0
0.001
0.002
0.003
0.004
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Red
uctio
n in
con
cent
ratio
n, m
g/L
L(ET)=0.5Lp(LEFT) L(ET)=0.5Lp(RIGHT) L(ET)=Lp
Figure 5.42. Reduction in solute concentration (after the source is removed) for different LET
lengths and locations.
0
5000
10000
15000
20000
25000
30000
35000
40000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Solu
te m
ass,
g
NA (Source removed) ET only NA only ET, Source removed
Figure 5.43. Solute mass in aquifer after removing the source, (a), and with a
phytoremediation system (b).
166
0
5
10
15
20
25
30
35
40
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
time, d
Sol
ute
mas
s, g
NA, Source ON L(ET)=0.5Lp, Left, Source ONNA(Source removed) L(ET)=0.5Lp, RightL(ET)=0.5Lp, Left L(ET)=Lp
% reduction in solute mass at different times
-10.0
0.0
10.0
20.0
30.0
40.0
50.0
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
% r
educ
tion
in s
olut
e m
ass
LpLeftRight
Figure 5.44. Solute mass reduction due to applying a phytoremediation system where the
contaminant source is removed.
5.7 Phytoremediation System Design Methodology
A set of design charts for estimating a phytoremediation system dimensions for different
remediation goals (of reducing the solute mass in the aquifer, reducing the plume length to a certain
value, and/or reducing the average solute mass-flux at a certain cross-section downstream the
contaminant source) and different TSCF values (according to the contaminant and tree types) are given
in Figure 5.45 through Figure 5.49. The set of design charts in this section investigates the effect of the
relative ET width to the source width, Ws for different values of TSCF on the design outcomes
explained at the end of section 5.2.
For the mass-removal design charts represented in Figure 5.45, the higher the ratio, s
ET
WW , the more
the solute mass removal. Figure 5.46 can be used in design purposes providing that the source width,
TSCF, and the reduction of the solute mass are known, so that the ET width can be estimated for a
certain RAO of solute mass reduction. In Figure 5.46, M* denote the solute mass in the aquifer at the
end of simulation period when a phytoremediation system is active, relative to M which represents the
solute mass in the aquifer under natural attenuation conditions only. Figure 5.46 shows that the higher
TSCF, the less the solute mass in aquifer (meaning more mass removal).
A similar series of design charts are produced for the two other design metrics including plume
length, and average contaminant mass-flux fore a wide variety of different modeling parameters. Figure
167
5.47 represents the effect of the relative phytoremediation system width, WET to the source width, Ws
on the plume length for different values of TSCF. Figure 5.48 represents the effect of WET/Ws values
on the average contaminant mass-flux, and Figure 5.49 represents the average mass-flux reduction vs.
TSCF for different values of (WET/Ws) and LET.
The design charts can be used to estimate the required phytoremediation width and length to
achieve a certain design goal included in the design metric. The design charts presented in this section
are also a useful decision making tool to decide if phytoremediation is the right option for the site
remediation. Two design examples for using the charts in designing a phytoremediation system for
plume length control are introduced in section 5.7.1 and 5.7.2.
168
Solute mass removal, L(ET)=Lp
0
2000
4000
6000
8000
10000
12000
100 150 200 250 300
W(ET), m
Solu
te m
ass
rem
oval
, g
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
Solute mass removal, L(ET)=0.5Lp
0
2000
4000
6000
8000
10000
12000
100 150 200 250 300
W(ET), m
Solu
te m
ass
rem
oval
, g
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
L(ET)=Lp
24
26
28
30
32
34
36
0 0.5 1 1.5 2 2.5 3
Thou
sand
s
W/Ws
Mas
s-in
aqu
ifer,
g
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
L(ET)=0.5Lp
24
26
28
30
32
34
36
0 0.5 1 1.5 2 2.5 3
Thou
sand
s
W/Ws
Mas
s-in
. gTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
L(ET)=Lp
0.7
0.75
0.8
0.85
0.9
0.95
1
0 0.5 1 1.5 2 2.5 3
W/Ws
Mas
s-in
/Mas
s-in
(NA)
TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
L(ET)=0.5Lp
0.7
0.75
0.8
0.85
0.9
0.95
1
0 0.5 1 1.5 2 2.5 3
W/Ws
Mas
s-in
/Mas
s-in
(NA)
TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
a) LET=Lp b) LET=0.5Lp
Figure 5.45. Effect of WET on solute mass removal for different TSCF values for a) LET=Lp,
and b) LET=0.5Lp.
169
L(ET)=Lp
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
0 0.2 0.4 0.6 0.8 1
TSCF
M*/M
NAW/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
L(ET)=0.5Lp
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
0 0.2 0.4 0.6 0.8 1
TSCF
M*/M
NAW/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
a) LET=Lp b) LET=0.5Lp
Figure 5.46. Effect of the TSCF on solute mass removal for different values of (WET/Ws) for
a) LET=Lp and b) LET=0.5Lp.
170
600
650
700
750
800
850
900
950
1000
100 150 200 250 300
ET Width, m
Plum
e Le
ngth
, m
NATSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
600
650
700
750
800
850
900
950
1000
100 150 200 250 300
ET Width, m
Plum
e Le
ngth
, m
NATSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
W/Ws
Lp*/L
p
TSCF=1.0TSCF=0.75
TSCF=0.50TSCF=0.25TSCF=0.00
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
W/Ws
Lp*/L
p
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
0 0.25 0.5 0.75 1
TSCF
Lp*/L
p
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
0 0.25 0.5 0.75 1
TSCF
Lp*/L
p
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
a) LET=Lp b) LET=0.50Lp
Figure 5.47. Design charts for the ET width required to reduce the plume length to a certain
design value for different TSCF values for a) LET=Lp and b) LET=0.5Lp.
171
X=500, L(ET)=Lp
0
500
1000
1500
2000
2500
3000
3500
4000
0 0.5 1 1.5 2 2.5 3
W(ET)/Ws
Av. M
ass-
flux,
mg/
d
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
X=500, L(ET)=0.5Lp
-500
0
500
1000
1500
2000
2500
3000
3500
4000
0 0.5 1 1.5 2 2.5 3
W/Ws
Av.
Mas
s-flu
x, m
g/d
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
X=1000, L(ET)=Lp
-50
0
50
100
150
200
250
300
350
400
450
0 0.5 1 1.5 2 2.5 3
W(ET)/Ws
Av.
Mas
s-flu
x, m
g/d
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
X=1000, L(ET)=0.5Lp
-50
0
50
100
150
200
250
300
350
400
450
0 0.5 1 1.5 2 2.5 3
W/Ws
Av. M
ass-
flux,
mg/
dTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
Figure 5.48. Effect of TSCF on average contaminant mass-flux for LET=Lp and LET=0.5Lp.
172
0
500
1000
1500
2000
2500
3000
0 0.2 0.4 0.6 0.8 1
TSCF
Av. M
ass-
flux,
mg/
d
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
-500
0
500
1000
1500
2000
2500
3000
3500
0 0.2 0.4 0.6 0.8 1
TSCF
Av. M
ass-
flux,
mg/
d
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
X=500 X=500
-50
0
50
100
150
200
250
0 0.2 0.4 0.6 0.8 1
TSCF
Av. M
ass-
flux,
mg/
d
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
-10
40
90
140
190
240
0 0.2 0.4 0.6 0.8 1
TSCF
Av.
Mas
s-flu
x, m
g/d
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
X=1000 X=1000 a) LET=Lp b) LET=0.5Lp
Figure 5.49. Effect of WET/Ws on average contaminant mass-flux for a) LET=Lp and b)
LET=0.5Lp.
173
5.7.1 Design Example 1
Design preliminary phytoremediation system to reduce the plume length from Lp to Lp* at the
compliance well, Figure 5.50, provided that the given parameters are:
• In-flow, m3/day • Ws (width of the source) • Lp (length of the dynamically stable plume) • In-flow = out-flow • TSCF=1.0
Source ComplianceWell
In-Flow
Lp*
Lp
Figure 5.50. Employing the design charts for a design problem.
Steps of the solution:
1- Assume the length of the phytoremediation area = Lp
2- Use the following chart (Figure 5.51) to find the width of the phytoremediation area.
For example, 70.0* =p
p
LL
, 62.2≅sW
W
The width of the plantation area (WET) = 2.62 the source width, Figure 5.51.
174
Calculating the density of trees:
Assuming the tree species is recorded to uptake 5 gal/day = 0.01892706 m3/d,
Max. ET rate = ( )ETAreaflow-In = ET rate per m2.
Number of trees in m2 = ET rate per m2/ ET rate of one tree
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
W/Ws
Lp*/L
p
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
Figure 5.51. Estimating the phytoremediation system width for a given reduction in plume
length.
175
5.7.2 Design Example 2
What would be the value for TSCF if the plume length was to be reduced 25% ( 75.0*
=p
p
LL
) using an
ET area with 6.2=sW
W . Using the following chart, and after drawing the 6.2=sW
W line by
interpolation, TSCF should be greater than or equal 0.54, Figure 5.52.
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
0 0.25 0.5 0.75 1
TSCF
LP*/L
P
W/Ws=3.0W/Ws=2.5W/Ws=2.0W/Ws=1.5W/Ws=1.0
0.54
W/Ws=2.6
Figure 5.52. Estimating the value of TSCF for a given phytoremediation system width to
reach a certain reduction in plume length.
176
C h a p t e r 6
Alternative Model for SEAM3D-PUP
6.1 Introduction
In this chapter, alternative models for plant uptake are investigated. In previous three chapters, the
original SEAM3D-PUP model is presented and tested based on the one model for plant uptake. This
may limit the flexibility of the model during calibration to data from controlled experiments or
remediation systems in the field.
6.1.1 Plant Uptake – Power Relationship
In the original SEAM3D-PUP model, the relationship between the solute concentration in
groundwater, C, and the concentration in the plant transpiration stream, CT, was assumed linear. This
approach is based on carefully-controlled experiments in the laboratory, which showed a linear trend
for C versus CT for a wide variety of solutes/plants (Shnoor 1995, 1997, 2002). Recent research at field
sites has shown a correlation between the amounts of contaminants observed in tree tissue and the
concentration of contaminants measured in the groundwater (Vroblesky et al., 1999; Ma and Burken
2002). Although experimental results from the laboratory show a linear trend for C versus CT in the
case of this particular solute (tetrachloroethene, PCE) and plant (hybrid poplar), the regression of the
field data suggests that this could be different for certain individual cases, as shown in Figures 6.1 and
6.2 (Struckhoff and Burken, 2005). The data points in there figures were extracted from the original
figures and are re-plotted. A regression analysis was performed to find the best fit equation, which was
, or on the form of 787.07552.0 XY ×= ( ) NT CTSCFC ×= , where N = empirical exponent. This
suggests that an alternative model for plant uptake may be more appropriate and less constraining to
simulate the contaminant fate in a phytoremediation system relative to the linear model.
177
μg PCE/L groundwater
μg P
CE/
kg tr
ee c
ore
0.10 1.0 10.0 100 1000 10,0000.10
1.0
10.0
100
1000
10,000
Figure 6.1. Relationship of PCE in tree cores collected at the New Haven Site plotted versus
the groundwater concentration below each tree at (6 – 7.6 m).
μg PCE/kg soil0.10 1.0 10.0 100 1000 10,000
g PC
E/kg
tree
cor
e
0.10
1.0
10.0
100
1000
10,000
100,000 1000,000
100,000
R=0.48
R=0.88
Figure 6.2. Relationship of PCE in tree cores collected at the New Haven Site plotted versus
the soil concentration 1.2 m below the surface near the base of the tree.
178
6.1.2 Plant Uptake – Plant Concentration Capacity
The plants that can uptake and accumulate toxic contaminants from groundwater without showing
symptoms of toxicity are called hyperaccumulators, (Baker and Brooks, 1989). Hyperaccumulators can
survive high contaminant concentrations compared to other plant species but still was not effective to
be used in phytoremediation because of their small sizes and slow growth rate, (Cunningham and Ow,
1996).
To investigate the effect of toxic contaminants on trees used in phytoremediation, Dietz (2000) and
Dietz and Schnoor (2001) tested a series of nine chlorinated aliphatic compounds (Table 6.1) for
phytotoxicity to hybrid poplar (Populus deltoids × Populus nigra ‘DN34’). Pre-rooted 20-centimeter (8-
inch) cuttings of hybrid poplar (Populus deltoids × Populus nigra ‘DN34’) were grown hydroponically with
the lower root portion in sealed reactors to minimize volatilization (Thompson et al. 1998).Chemical
solutions were replaced every 2 days to maintain a constant exposure concentration.
Phytotoxicity tests were conducted with triplicate reactors dosed for a period of 2 weeks. At the
higher solvent concentrations (above the zero growth levels), wilting of shoots and damage to roots
were observed. At concentrations between zero-growth and hall zero-growth levels, fine root
formation was arrested, similar to other studies (Newman et al. 1997).
Reduction in total biomass and transpiration were monitored as indicators of acute toxcicity, and
both showed similar patterns. Highly chlorinated aliphatic compounds were more toxic to poplar
(Populus spp.) cuttings than compounds with fewer chlorine atoms within the set of five ethenes or
four ethanes tested (Table 6.1). The ethenes were more toxic than the corresponding ethanes,
contrasting with, results by Schubert et al. (1995).
179
Table 6.1. Toxic Effects on Hybrid Poplar (Populus deltoides × Populus nigra DN34) from Chlorinated Aliphatic Compounds (Dietz and Schnoor 2001).
Chemical log Kow Zero-growth concentration, mg/L
50 percent transpiration concentration, mg/L
Tetra-chloro-ethylene 3.4 45 ± 3 38 ± 6 Tri-chloro-ethylene 2.42 118 ± 12 131 ± 22 trans-Dichloro-ethylene 2.06 465 ± 50 349 ± 74 cis-Dichloro-ethylene 1.86 582 ± 57 494 ± 83 1,1-Dichloro-ethylene 2.13 543 ± 54 281 ± 56 1,1,2,2-Tetra-chloro-ethane 2.39 151 ± 17 151 ± 34 1,1,2-Ttichloroethane 2.07 307 ± 20 253 ± 36 1,1,1-Tti-chloro-ethane 2.49 267 ± 29 160 ± 33 1,1-Dichloro-ethane 1.79 1059 ± 109 802 ± 165 "Source Hoard, P.H., ed. (1990) Handbook of Environmental Fate and Exposure Data for Organic Chemicals. Lewis Publishers, Chelsea, Michigan U.S.
The previous study suggests that the plant has a specific maximum tolerance to the contaminants
due to toxicity. Bulk flow in the xylem from root to shoot is driven by transpiration from the shoot,
which creates a negative pressure in the xylem that pulls up water and solutes (Taiz L and Zeiger E.,
2002). Species such as poplar are phreatophytes, or water spenders; they have long roots that tap into
the ground water. Mature poplar trees can transpire 200–1000 liters of water per day (EPA, 1999;
Wullschleger et al., 1998). In addition to plant species composition, vegetation height and density affect
transpiration, as well as environmental conditions: Transpiration is generally maximal at high
temperature, moderate wind, low relative air humidity, and high light (Taiz L and Zeiger E., 2002).
Consequently, phytoremediation mechanisms that rely on translocation and volatilization are most
effective in climates with low relative humidity and high evapotranspiration.
6.1.3 Objective
The present chapter describes alternative models for plant uptake that incorporate non-linear,
equilibrium relationship between contaminant concentrations in the saturated zone and plant
transpiration stream. Two new models for plant uptake are proposed; 1) one based on the power
function observed in field data, and 2) one designed to investigate plant tolerance to VOC
contaminants in groundwater by assuming that the tolerance of a plant to a VOC is reflected in the
relationship between the contaminant concentrations in groundwater and the maximum VOC
180
concentration in plant tissue. Both models are described and implemented in SEAM3D-PUP. Results
from both models are compared with the linear model.
6.2 Mathematical Models
The mathematical models for plant uptake are analogous to those for sorption of a hydrophobic
contaminant in soil and aquifer sediment. All three models describe partitioning between the aqueous
(groundwater) and transpiration (plant) phases and are based on the assumption of instantaneous,
equilibrium kinetics. In addition to the linear isotherm (TSCF model), two non-linear models are
considered:
1- Linear sorption isotherm
2- Freundlich sorption isotherm (Power function)
3- Langmuir sorption isotherm (Plant total concentration capacity)
6.2.1 Freundlich Isotherm (Power Function)
The Freundlich isotherm is a more general equilibrium sorption equation than the linear equilibrium
model. It was developed mainly to allow for an empirical account of the variation in adsorption heat
with concentration of an adsorbate (vapor or solute) on an energetically heterogeneous surface (Chiou
2002). When an adsorption relationship can be plotted as a straight line on log-log paper, it is described
by the Freundlich isotherm. For plant uptake, the relationship between a solute concentration of the
species i in groundwater, C, and the concentration in the plant transpiration stream, CT, is expressed by
a power function
………. (6.1)
where K is a coefficient equal to TSCF at C = 1 [L3 Mi] and N is an empirically-based exponent.
Ni
Ti KCC =
The slope of the curve on a log-log plot of C versus CT is represented by N. In sorption studies, the
N value is in principle less than 1, because the adsorption isotherm is commonly concave to the C axis,
and varies with the extent of adsorption. Depending on the adsorbent, the constancy of N may apply
to a narrow or wide range of C. In the case of plant uptake, it can be determined from the slope of the
plot of log CT versus log C over a specific range.
181
The limitations of Freundlich sorption isotherm includes (Zheng & Bennett, 1995):
• Assumes that there is a unlimited number of available sorption sites.
• Validity is limited to the limits of experimentally derived data.
Using mass balance of the solute in groundwater and the Freundlich sorption isotherm to drive the
governing equation for plant uptake
( ) ETNi
i qCTSCFt
C−=
∂∂θ ………. (6.2)
which can be substituted for Equation (3.4).
6.2.2 Langmuir Sorption Isotherm (Plant Tolerance)
Langmuir (1918) considered the adsorption of gases or vapors on a plane surface that contains a
fixed number of identical active sites. From a kinetic consideration, the rate of vapor desorption from
the occupied sites is set equal to the rate of adsorption on the unoccupied sites at equilibrium. In the
case of sorption experiments, the ratio of the aqueous to solid-phase concentrations (C/C*) are plotted
versus C on arithmetic graph paper. If this falls on a straight line (Figure 6.3), it is the nonlinear
Langmuir adsorption isotherm (Olsen & Watanabe 1957), which is of the form
CC
Cα
αβ+
=1
* ………. (6.3)
where β is the concentration of sorption sites or the maximum sorption capacity and α is the Langmuir
constant.
182
C (M/L3)
C* (M/M
)
C (M/L3)
C/C
* (M/M
)
αβ1
β1
1
Curvilinear on linear graph paper Straight line on log-log graph paper,
CCC
βαβ11
* +=
Figure 6.3. The Langmuir nonlinear equilibrium isotherm.
The Langmuir isotherm can be adapted for the case of plant uptake to account for plant tolerance
to a VOC or semi-volatile organic compound as follows
i
icTi CK
CTKC1
1
1+×
= ………. (6.4)
where K1 ⎟⎟⎠
⎞⎜⎜⎝
⎛ML3
, and Tc ⎟⎠⎞
⎜⎝⎛
3LM are constants dependent on the compound and the susceptibility of the
plant to toxicity effects. At low concentrations where K1C << 1, the model is linear where K1×Tc =
TSCF. At relatively large groundwater concentrations where K1C >> 1, the model reaches a constant
value where . cTi TC =
The expression for mass loss due to plant uptake becomes
ETi
ici qCKCTK
tC
⎟⎟⎠
⎞⎜⎜⎝
⎛+×
−=∂
∂
1
1
1θ ………. (6.5)
which can be substituted for Equation (3.4).
183
6.3 Model Verification
The same test case used to verify the original SEAM3D-PUP code (Figure 4.1 – Closed system
model with single stress period) was used to verify the new alternative SEAM3D-PUP code. For each
new option (Freundlich, ISO=2, and Langmuir, ISO=3), the output concentrations and solute mass
uptaken are calculated manually at different time steps. The manually calculated results are then
compared with the alternative code output.
6.3.1 Freundlich (ISO=2) Verification
The solute mass removed from the model due to the trees sink effect was calculated at different
time steps using the equation: ( )( ) tCTSCFQMMMM NEToocalc Δ−=Δ−= , where M0 is the starting
mass time = 0 and QET is the evapotranspiration rate which is a function of surface elevation,
maximum ET rate, and root extinction depth. The QET value is set to be maximum in this model
simulation, C is the solute concentration in groundwater, N is the Freundlich power constant (set to be
equal to 2.0), and TSCF is set equal to 1.0. Once mass removal is calculated, the new solute
concentration in groundwater at the end of time step is calculated using the equation
∑=
Δ−=
Δ−=Δ−= n
i
calco
fluid
calcoocalc
Ah
MCnV
MCCCC
1θ
where A is the total model area, h is the average
hydraulic head, and θ is the effective porosity. The manual calculations for C and M are presented in
Table 6.2 and the SEAM3D-PUP results for concentration and mass are listed in Table 6.3. The
manual calculations and the SEAM3D-PUP results showed perfect match, indicating the code was
correctly formulated. The sensitivity of the SEAM3D-PUP results (mass and aqueous concentration)
to different values of N are listed in Tables 6.4 and 6.5, respectively, and are shown in Figure 6.4 and
6.5, respectively.
184
Table 6.2. Manual calculations of concentration and mass using manual calculations based on the Freundlich model for the closed system test case.
Time step, d C, mg/L ΔM, g Total M, g
2 9.5 -20,000 -20,000 4 9.04875 -18,050 -38,050 6 8.639 -16,376 -54,426 8 8.266 -14,927.7 -69,353.7 10 7.924 -13,665.9 -83,019.5
Table 6.3. Mass, mass removal, and concentration results using the SEAM3D-PUP Freundlich model for plant uptake for the closed system test case.
TIME TOTAL IN
TOTAL OUT SOURCES SINKS
NET MASS FROM FLUID-STORAGE
TOTAL MASS IN AQUIFER
LOCATION OF OBSERVATION POINTS (K,I,J) = (1,1,1)
(d) (g) (g) (g) (g) 2 20000 -20000 0 -20000 0 380000 9.50004 38050 -38050 0 -38050 0 362000 9.04876 54426 -54426 0 -54426 0 346000 8.63948 69354 -69354 0 -69354 0 331000 8.2662
10 83020 -83020 0 -83020 0 317000 7.9245
Table 6.4. Mass removal for the closed-system test case using the SEAM3D-PUP Freundlich model for plant uptake for different values of (N).
Iso=2, N Time step, d 1 0.75 0.5 0.25 0
0 0 0 0 0 02 2000 1124.7 632.46 355.66 2004 3990 2247 1264.4 711.23 4006 5970.1 3366.9 1895.9 1066.7 6008 7940.2 4484.5 2526.8 1422.1 800
10 9900.5 5599.7 3157.3 1777.5 1000
Table 6.5. Solute concentration in groundwater for the closed-system test case using the SEAM3D-PUP Freundlich model for plant uptake for different values of (N).
Iso=2, N Time step, d 1.0 0.75 0.50 0.25 0.0
0 10 10 10 10 102 9.95 9.9719 9.9842 9.9911 9.9954 9.9002 9.9438 9.9684 9.9822 9.996 9.8507 9.9158 9.9526 9.9733 9.9858 9.8015 9.8879 9.9368 9.9644 9.98
10 9.7525 9.86 9.9211 9.9556 9.975
185
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
0 2 4 6 8 10
Time, d
Mas
s O
ut (S
inks
), g N=1.0
N=0.75N=0.5N=0.25N=0.0Iso=1
(a)
9.7
9.75
9.8
9.85
9.9
9.95
10
0 2 4 6 8 10
Time, d
Conc
., m
g/L
N=1.0N=0.75N=0.5N=0.25N=0.0Iso=1
(b)
Figure 6.4. SEAM3D-PUP results for ISO=2 for a) Solute mass removal, and b) solute
concentration.
186
0
1000
2000
3000
4000
5000
6000
0 2 4 6 8 10
Time, d
Mas
s ou
t (S
inks
), g
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
(a)
9.84
9.86
9.88
9.9
9.92
9.94
9.96
9.98
10
10.02
0 2 4 6 8 10
Time, d
Con
c., m
g/L TSCF=1.0
TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
(b)
Figure 6.5. Effect of TSCF using ISO-2 for a) Solute mass removal, and b) solute
concentration for initial source concentration = 10 mg/L, and N=0.75.
187
To further elucidate the isotherm trends (of linear, power, and maximum capacity), the problem was
re-run for different values of TSCF and ISO-2 power constant (N) and a range of starting
concentration. The results are summarized in Figure 6.6. The plots of groundwater concentration
versus solute mass for different values of TSCF are similar to Figure 6.1. As expected, the higher the
value of TSCF, the greater the solute mass loss at the end of the simulation. In Figure 6.6.b., the
higher the ISO-2 power constant, the more solute mass is removed for the same initial concentration.
6.3.2 Langmuir (ISO=3) Verification
In this case, solute mass removed from the model due to the trees sink effect was calculated as a
function of time using the equation MMM ocalc Δ−= , where M0 is the starting mass at time = 0 and
ΔM is the mass removal by trees at the end of time step Δt and can be calculated from the equation:
tCCQ
CKTK
Vn ETc
te ∂∂
=+
×−
1
1
11 , which can be approximated to tQ
CKTKM ET
c Δ×+
×−=Δ
1
1
1. Once the
mass removed is calculated, the new solute concentration in groundwater at the end of time step is
calculated using the equation ∑
=
Δ−=
Δ−=Δ−= n
i
calco
fluid
calcoocalc
Ah
MCnV
MCCCC
1θ
where A is the total
model area, h is the average hydraulic head, and θ is the effective porosity. The manual calculations for
C and M and SEAM3D are represented in Tables 6.6 and 6.7, respectively for the case where K1 = 0.8
and Tc = 8.0. Again, a comparison of the manual and SEAM3D results showed an identical match.
Table 6.6. Manual calculations of concentration and mass using manual calculations based on the Langmuir model for the closed system test case.
Time step, d C, mg/L ΔM, g Total M, g
1 9.955 -1777.78 -1777.78 2 9.911 -1776.89 -3554.67 3 9.866733 -1776.008406 -5330.68 4 9.822355 -1775.113784 -7105.8 5 9.778 -1774.212444 -8880
188
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
0 5 10 15 20
C, mg/L
Mas
s (S
inks
), g
TSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25
(a) N=0.75
1000
10000
100000
1 10 100
C, mg/L
Mas
s (S
inks
), g
N=1.0N=0.75N=0.25
(b) TSCF=1.0 (log-log scale)
Figure 6.6. Effect of starting concentration on mass removal using ISO-2 modeling option
for a) N=0.75 and different values of TSCF, and b) TSCF=1.0 and different values of N.
189
Table 6.7. Mass, mass removal, and concentration results using the SEAM3D-PUP Langmuir model for plant uptake for the closed system test case.
TIME TOTAL IN
TOTAL OUT
SOURCES SINKS
NET MASS FROM FLUID-STORAGE
TOTAL MASS IN AQUIFER
LOCATION OF OBSERVATION POINTS (K,I,J) = (1,1,1)
(d) (g) (g) (g) (g) g 2 1777.8 -1777.8 0.0000 -1777.8 0.0000 398222 9.9556 4 3554.6 -3554.7 0.0000 -3554.7 0.0000 396446 9.9111 6 5330.7 -5330.7 0.0000 -5330.7 0.0000 394669 9.8667 8 7105.8 -7105.8 0.0000 -7105.8 0.0000 392894 9.8224 10 8880.0 -8880.0 0.0000 -8880.0 0.0000 391120 9.7780
For the ISO-3 simulation option, two variables have to be estimated first. The Langmuir plant
uptake constant, K1 (L3/M), and the total plant concentration capacity, Tc (M/L3). The selection of the
two parameters K1 and Tc depends on the fitted field or lab data for the solute concentration in
groundwater, C versus the solute mass concentration in the plant (represented in terms of total solute
mass in plant/plant core mass).
Assuming different values for K1, the SEAM3D-PUP simulation using ISO-3 option, indicated that
the lower the value of K1, the lower the solute concentration, Figure 6.7a, and the higher the solute
mass removal, Figure 6.7b.
190
8
8.2
8.4
8.6
8.8
9
9.2
9.4
9.6
9.8
10
0 2 4 6 8 10
Time, d
Con
c., m
g/L K1=1.0
K1=0.75K1=0.50K1=0.25K1=0.0
(a)
0
10
20
30
40
50
60
70
80
0 2 4 6 8 10
Thou
sand
s
Time, d
Mas
s ou
t, g
K1=1.0K1=0.75K1=0.50K1=0.25K1=0.0
(b)
Figure 6.7. Concentration (a), and solute mass removal (b) vs. time for different values of
ISO-3 constant, K1 (Tc=8.0).
191
As the plant total concentration capacity, Tc, increases, the ability of the plant to translocate more
solute mass from groundwater increases. Table 6.8 lists the results of solute mass concentration in
groundwater versus mass removal by plants for different values of Tc. The graphical display of the
results is shown in Figure 6.8.
To show the trend or relationship between the solute concentration in groundwater, C, and the
solute mass uptaken by a phytoremediation system, the three simulation options are plotted in Figure
6.9. The plotted values are following the trends expected for the linear sorption isotherm, Freundlich
sorption isotherm (Power function), and Langmuir sorption isotherm (Plant total concentration
capacity).
Table 6.8. Solute concentration at the end of the simulation and solute mass loss for different plant total concentration, Tc.
Tc=5.0 Tc=8 Tc=10.0 Initial Conc., mg/L C M C M C M
0 0 0 0 0 0 01 0.92975 2810 0.8619 5523.9 0.88875 4450.1
2.5 2.392 4321.2 2.2853 8587.4 2.3278 6887.75 4.8687 5251.4 4.738 10479 4.7902 8390.66 5.8639 5445.4 5.7282 10872 5.7824 8703.87 6.8602 5592.8 6.7207 11171 6.7765 8941.48 7.8573 5708.4 7.7149 11405 7.7718 9127.79 8.855 5801.6 8.7102 11593 8.7681 9277.8
10 9.853 5878.2 9.7063 11748 9.765 9401.220 19.844 6248.8 19.688 12495 19.75 9996.950 49.838 6493.3 49.675 12986 49.74 10389
192
Langmuir, Iso=3, K1=0.75
0
2000
4000
6000
8000
10000
12000
14000
0 5 10 15 20
Cw, mg/L
Mas
s (S
inks
), g
Tc=10.0Tc=8.0Tc=5.0
Figure 6.8. Effect of plant total concentration capacity, Tc on solute mass removal for ISO-3.
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20
Thou
sand
s
Cw, mg/L
Mas
s (S
inks
), g
Linear, Iso=1 Freundlich, Iso=2 Langmuir, Iso=3
Figure 6.9. Comparing the three different Isotherms.
193
6.4 Alternative Model Applications, PCE simulation
One of the several model runs in chapter 4 is selected to demonstrate the different SEAM3D-PUP
alternative model options. The model selected is described in (Figure 5.2). First, the results are verified
by setting a simulation for ISO=2 and set the power constant to 1.0 (which will reduce to linear C/M
relationship). The TSCF value is set to be equal to 0.7552 to resemble the PCE data, (Struckhoff and
Burken, 2005). The results of solute mass change versus time and solute concentration versus distance
of the original and alternative SEAM3D-PUP are plotted in Figure 6.10. The results for both the
original and modified SEAM3D-PUP showed perfect match.
The recorded field and lab results for PCE uptake showed that the best simulation option is by
using isotherm=2 with TSFC= 0.7552 and N = 0.787 according to the fitting equation
, which is on the form 787.07552.0 XY ×= ( ) ( )NT CTSCFC ×= . The simulation of PCE uptake using
ISO=3 option will depend on the plant maximum concentration capacity (Tc). This factor can be
assumed equal to 0.0 to 0.8, however, the ISO=3 simulation option is best suiting simulating site with
high solute concentration in groundwater (Table 6.1). The Tc value will be assumed to be equal to 0.6.
The higher the value of Tc, the closer the results to the linear simulation option (ISO=1).
Figures 6.1 and 6.12 can be used to estimate mass removal, and concentration profile specific PCE
with TSCF=0.7552 and N = 0.787.
194
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s, g ISO=2, N=1.0
Linear
(a) Solute mass removal of PCE
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
ISO=2, N=1.0Linear
(b) PCE Concentration profile
Figure 6.10. Comparing SEAM3D-PUP alternative model with ISO=2, and N=1.0 and the
linear original code.
195
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s (in
aqu
ifer)
, g
ISO-1 ISO-2 ISO-3, Tc=0.6 ISO-3, Tc=0.7
(a)
0
20
40
60
80
100
120
140
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s (S
ink)
, g
ISO-1 ISO-2 ISO-3, Tc=0.6 ISO-3, Tc=0.7
(b)
Figure 6.11. Mass-in aquifer (a), and solute mass removal (sinks) (b) for PCE with
TSCF=0.7552 and N = 0.787.
196
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist, m
Con
c., m
g/L
ISO-1 ISO-2 ISO-3, Tc=0.6 ISO-3, Tc=0.7
Figure 6.12. Concentration profile for PCE.
197
C h a p t e r 7
Conclusions and Recommendations
A general groundwater solute transport with phytoremediation model was developed to study fate
and movement of organics in the presence of vegetation. The model consisted of two components:
one for root-sorption and one for plant uptake. The mathematical model was solved with a finite
difference-based algorithm using the original SEAM3D code and adding a new separate Plant Uptake
Package (SEAM3D-PUP). The code was verified by comparing the results of root sorption from
SEAM3D-PUP to the results of SEAM3D/MT3DMS Reaction Package for sorption to the aquifer
matrix. For the plant uptake problem, the results of SEAM3D-PUP were verified through comparison
to the results of the SEAM3D/MT3DMS Source-Sink Mixing Package.
This study demonstrates the usefulness of numerical groundwater modeling in addressing several
issues pertaining to the design or evaluation of a phytoremediation system which depends on
phreatophytes. While the direct uptake or translocation of contaminants is not explicitly addressed, the
engineered system of deep-rooted poplars trees (or similar species) was predicted to provide a large
degree of hydraulic control, despite seasonal variation in water use rates by the plantation. The
evapotranspiration was turned periodically on and off to simulate seasonal changes in plants
consumption.
Modeling clearly has application at phytoremediation sites for evaluating or designing a containment
system with respect to factors such as tree planting density (by changing the maximum ET rate), plume
width versus groundwater flow rate, seasonal effects, residence time of groundwater within the
microbially active rhizosphere, prediction of downgradient distance where the contaminant
concentration reaches a point of compliance (POC), and future modifications to the system design to
reduce the contaminate mass-flux even after the contaminant source is removed.
The model was used to investigate several site design parameters for phytoremediation including
plantation width (WET) and length (LET), groundwater flux compared to ET flux, and the effect of
using phytoremediation after the contamination source is removed. Each of those parameters was
198
tested with respect to three output metrics: solute mass removal and decreases in solute concentrations
and solute mass-flux.
In general, modeling researches on phytoremediation helped to determine the various mechanisms
involved in movement of soil constituents in presence of plants. This model could also be utilized in
design to predict the feasibility of using trees/phytoremediation for controlling or remediation
contaminated soils and groundwater. Phytoremediation is economically competitive and results are
impressive to regulators and user communities. Enhanced biodegradation in presence of plants occurs
in this process but was not demonstrated in this study because the focus of new model was plant
uptake and root sorption. SEAM3D has a Biodegradation Package which can be used to simulate the
rhizosphere biodegradation effect. The root zone supports an eutrophic environment by exuding
sloughed root masses and rhizodeposits that provide carbon and energy to diverse microbial consortia
indigenous to soil.
The alternative model presented in Chapter 6 extended the capabilities of plant uptake simulation to
include three different ways of addressing the concentration/mass relationship. The first approach
represented in the original SEAM3D-PUP assumes a linear relationship between groundwater solute
concentrations and relative the transpiration stream concentrations, which is analogous to linear
isotherm sorption. The linear approach can be used in situations of low groundwater solute
concentrations because it is assumed the plants are capable of transpiring the whole solute mass
without subjecting to toxicity. The alternative model includes two other concentration relationships.
The first is the non-linear power function analogous to Freundlich isotherm. That approach (which is
referred to as ISO=2) is suitable for a larger range of moderate solute concentrations. The second
model is analogous to the Langmuir isotherm and suggests that the plant reaches a maximum capacity
of handling (or transpiring) solute mass due to high solute concentrations in groundwater. The high
solute concentration may lead to plant toxicity as suggested by (Dietz and Schnoor 2001). This
approach (referred to as ISO=3) is suitable for modeling phytoremediation systems in contaminated
sites with high solute concentrations.
The alternative models in the SEAM3D-PUP make use of recent field and laboratory finding for
different plant uptake measures. The statistical analysis of the groundwater/transpiration concentration
relationship will determine which trend is more appropriate for a given contaminant/plant
combination (i.e., linear, power, or power with a maximum capacity).
199
Researchers recently presented the results of a field investigation at a PCE-contaminated site
indicating that the experimental results suggests using the ISO-2 option with a value of TSCF = 0.7552
and a = 0.787 (Struckhoff and Burken, 2005). Other site conditions such as plume source
concentrations may suggest using different simulation options, but in all cases, field or lab records of
plant uptake are important to estimate RCF and TSCF values.
Recommendations for Future Research
1- Model improvements
SEAM3D reads the head values using the results from MODFLOW and uses the results from the
ET package which assumes a linear relationship between head and ET rate. A more comprehended ET
package that assumes a segmental relationship between h & ET rate may be applied, Figure (7.1)
Hyd
raul
ic H
ead,
L
QET(L3/T)
ETSX
(ext
enct
ion
dept
h)ET
Q
Max
imum
Evap
otra
nspi
ratio
n
h
A
B
C
D
Segment 1 Segment 2Segm
ent 3
h
h
d
(h - d)s
QET
M
ETQ
hsh
Max
imum
Evap
otra
nspi
ratio
n
Land surface elevation (SURF)
d
Slop
e =
dETM
Q ____
__
0
s
Figure 7.1 Linear and segmental ET packages.
2- The RCF and TSCF are input into SEAM3D-PUP as design parameters, where the user has to
know these two values for different contaminants. There would be another alternative to the
previous procedure where the model can have a database of different types of contaminants
200
where the user may select the type of solute, and the software can estimate the RCF, and TSCF
form the database.
3- The software package needs a graphical user interface instead of editing ASCII text files for
the inputs. The GUI will be integrated into the original SEAM3D SI.
4- The software package needs to be tested against suitable field data.
5- SEAM3D has a separate biodegradation package which should be suitable to be used with
SEAM3D-PUP package to simulate the biodegradation rhizosphere effect.
6- There are good potential for conducting more statistical analysis and/or regression for the
results of the studied cases to come up with empirical relationships between the
phytoremediation system design parameters (including ET dimensions, location, ET rate, and
ET flux) and the site remediation goals (including solute mass reduction, downstream plume
concentration, and solute mass-flux) which can be easily used as a decision supporting tool.
7- The alternative model gives more flexibility for the designer/decision maker to select from
three different options according to the site situations (mainly the source concentration). The
selection is categorized according to the source concentration as follows:
a. Low source concentration: options 1 (Linear Isotherm), and option 2 (Freundlich
Isotherm) gives conservative results with respect to mass removal.
b. Medium source concentration: option 2 (Freundlich Isotherm) gives conservative
results with respect to mass removal.
c. High source concentration: option 3 (Langmuir Isotherm) gives conservative results
with respect to mass removal.
8- As mentioned in #7, it is totally up to the phytoremediation system designer to choose from
three different code options. The suitable code for each source concentration (as suggested by
the author in #7) needs more verification using recorded site data.
201
C h a p t e r 8
Input Instructions
SEAM3D MODEL INPUT
General Information
Estimation of model parameters for biodegradation may be based on laboratory measurements,
published values, and theoretical estimates. To produce maximum flexibility, SEAM3D allows
parameters to vary across the aquifer layers and among the various substrates and electron acceptors
for biodegradation. However, in the absence of detailed information, the user is advised to enter
identical parameter values to describe the layers and certain biodegradation processes. Thus, parameter
estimation can be simplified when available data do not support a more detailed analysis.
Types of Input
Like MT3DMS, input for SEAM3D may be formatted, list-directed, or unformatted.
Formatted
Input variables may be formatted as integer, real, character, or logical. In the detailed input
instructions (Sections 4.2.1 to 4.2.6), the format column uses I to specify an integer, F for a real
number, A for a character variable, and L for a logical variable. Input conventions follow the standards
of the FORTRAN 77 language.
List Directed
List directed, or free format, input involves a sequence of values separated by blanks or commas.
The list directed record terminates when a slash (/) is encountered, repeat counters are permitted, and
each new record should begin on a new line of the input file.
202
Unformatted
Unformatted files contain binary characters and must be written and read by the computer. Relative
to formatted files, unformatted files are smaller and can be processed more readily.
Array Readers
Most of the input data for SEAM3D is handled by the subroutines IARRAY and RARRAY in the
utility module of the program. IARRAY reads one or two dimensional integer arrays, and RARRAY
reads one or two dimensional real arrays. Three dimensional arrays are handled by reading a two
dimensional areal array for each model layer. Each time an array reader is called, it initially reads an
array control record, which occupies a single line of the input filed and is formatted as follows:
Record: IREAD CNSTNT (real) or FMTIN IPRN
ICONST (integer)
Format: I10 F10.0 (real) or A20 I10
I10 (integer)
If IREAD = 0, then RARRAY sets all elements of the array equal to CNSTNT, or IARRAY sets all
elements equal to ICONST.
If IREAD = 100, then array values (entered on the lines following the array control record) are read
in the format specified by FMTIN.
If IREAD = 101, then array values are read as blocks, which are entered on the lines following the
array control record. The first line contains only the record NBLOCK, which is an integer specifying
the number of blocks to follow. Each block occupies a single line, consisting of I1, I2, J1, J2, VALUE;
where I1 is the index of the first row of the block, I2 is the index of the last row, J1 is the index of the
first column of the block, J2 is the index of the last column, and VALUE is the value assigned to array
elements within the block.
If IREAD = 102, then array values are read as zones.
If IREAD = 103, then array values are read in list directed format.
203
If IREAD is equal to a nonzero value other than 100, 101, 102, or 103, then array values are read
from a separate file. If IREAD is positive, then IREAD is the unit number for the separate file, which
is formatted according to FMTIN. If IREAD is negative, then the separate file is unformatted, and the
absolute value of IREAD is its unit number.
If IREAD ≠ 0 and CNSTNT or ICONST ≠ 0, then all elements in the array are multiplied by
CNSTNT or ICONST.
The format specifier FMTIN must be enclosed in parentheses.
If IREAD ≠ 0, then IPRN acts as a flag to indicate whether the array will be printed for checking.
The array will not be printed in IPRN is negative.
Units
Like MT3DMS, SEAM3D requires the user to specify units and use consistent units for all input
and output variables. In addition, the time unit must be consistent with that used in the flow model.
The single exception to this rule involves the concentrations of solid phase electron acceptors, which
are entered as mass of electron acceptor per 1 x 106 mass of soil solids (e.g. micrograms per gram).
Units of METERS for length and GRAMS for mass are convenient because they produce
concentration units of grams per cubic meter, which is equivalent to milligrams per liter.
Input Instructions
Note: Input instructions for the extended alternative model are presented in red.
Input Instructions for the Plant Uptake Transport Package
This input file must be created only if the Phytoremediation Package is specified in the Basic
Transport Package; i.e., TRNOPT(10) is set to “T”. Input is read on unit 13, which is preset in the
main program. Input to the Phytoremediation (PUP) Package is read from the file that is type "PUP".
All non-array parameters are free format if the word FREE is specified in item 4 of the Basic Package
input file; otherwise, the non-array parameters have 10-character fields.
1. Record: FRCF FTSCF ISOTHMP Format: 2L2, I10
o FRCF is a logical flag for simulating root sorption;
204
o FTSCF is a logical flag for simulating transpiration. o ISOTHMP is a flag indicating which type of plant uptake is simulated:
ISOTHMP =1, Linear isotherm (equilibrium-controlled); =2, Freundlich isotherm (equilibrium-controlled); =3, Langmuir isotherm (equilibrium-controlled);
(Enter 2 if FRCF=T) 2. Array: RHOBR(NCOL,NROW)(one array for each layer) Reader: RARRAY
o RHOBR is the bulk density of the root medium (unit: ML-3 ). (Enter 3 for each species if ISOTHMP>1) 3. Array: SP1P(NCOL,NROW) (one array for each layer) Reader: RARRAY
SP1P is the first plant uptake parameter. The use of SP1P depends on the type of plant uptake selected (i.e., the value of ISOTHMP): For Freundlich plant uptake (ISOTHMP=2), SP1P is the Freundlich exponent N.
For Langmuir plant uptake (ISOTHMP=3), SP1P is the Langmuir equilibrium constant (Kl) (unit: L3 M-1).
(Enter 4 for each species if ISOTHMP>2) 4. Array: SP2P(NCOL,NROW) (one array for each layer) Reader: RARRAY
SP2P is the second plant uptake parameter. The use of SP2P depends on the type of plant uptake:
For Langmuir plant uptake (ISOTHMP=3), SP2P is the total concentration of the plant uptake sites available (Tc) (unit: ML-3).
FOR EACH STRESS PERIOD (Enter 5 through 8 if FRCF=T) 5. Record: INSURF INEXDP INRCF Format: 3I10
o INSURF--is the PUP/ET surface (SURF) read flag. If INSURF >= 0, an array containing the PUP/ET surface elevation (SURF)
will be read. If INSURF < 0, the PUP/ET surface from the preceding stress period will be
reused. o INEXDP--is the extinction depth (EXDP) read flag.
If INEXDP >= 0, an array containing the extinction depth (EXDP) will be read.
If INEXDP < 0, the extinction depth from the preceding stress period will be reused.
o INRCF--is the root concentration factor (RCF) read flag.
205
If INRCF >= 0, an array containing the root concentration factor (RCF) will be read for each species.
If INRCF < 0, the root concentration factors from the preceding stress period will be reused.
(Enter 6 if INSURF >= 0) 6. Array: SURF(NCOL,NROW)(one array for each layer) Reader: RARRAY
o SURF--is the elevation of the PUP/ET surface. (Enter 7 if INEXDP >= 0) 7. Array: EXDP(NCOL,NROW)(one array for each layer) Reader: RARRAY
o EXDP--is the PUP/ET root extinction depth. (Enter 8 for each species if INRCF >=0) 8. Array: RCF(NCOL,NROW) Reader: RARRAY
o RCF--is the root concentration factor. (Enter 9 through 10 if FTSCF=T) 9. Record: INTSCF Format: 1I10
o INTSCF--is the transpiration concentration factor (TSCF) read flag. If INTSCF >= 0, an array containing the transpiration concentration factor
(TSCF) will be read for each species. If INTSCF < 0, the transpiration concentration factors from the preceding
stress period will be reused. (Enter 10 for each species if INRCF >=0) 10. Array: TSCF(NCOL,NROW) Reader: RARRAY
o TSCF--is the transpiration concentration factor. o Enter TSCF=1.0 in case of ISO-3 because TSCF is implicitly simulated (K1×TC)
END INPUT
RCF Notes
• RCF must be between 0.0 and 1.0. Thus if the RCF is specified as less than 0.0 it will be set to
0.0 by the program. Correspondingly, if the RCF specified is greater than 1.0 it will be set to
1.0 by the program.
206
• PUP/ET parameters (PUP/ET surface and root extinction depth) are specified in two-
dimensional arrays, SURF and EXDP, with one value for each vertical column. Accordingly,
PUP/ET is calculated for one cell in each vertical column. IEVT (the layer indicator array code
from the EVT package) determines for which cell/layer in the column PUP/ET will be
calculated (See EVT input file).
• If root sorption is simulated, the first retardation factors displayed in the standard output file
represent retardation values without the root sorption effect. Following retardation factor
arrays represent the total values (including root sorption) recalculated for each stress period.
TSCF Notes
• TSCF must be between 0.0 and 1.0. Thus if the TSCF is specified as less than 0.0 it will be set
to 0.0 by the program. Correspondingly, if the TSCF specified is greater than 1.0 it will be set
to 1.0 by the program.
• If the Source/Sink Mixing Package (SSM) is also used, TSCF must be coordinated with CEVT
from SSM. Either TSCF or CEVT must be set to zero for all grid locations where EVTR is
less than or equal to zero (i.e. where EVTR indicates that water is exiting the model). If this is
not done, the same mass may be removed from the model domain twice and mass balance
errors will occur.
• TSCF only has meaning if the evapotranspirative water flux (EVTR) is negative, indicating that
water is being drawn out of the aquifer. If EVTR is positive for a cell with non-zero TSCF, the
program will terminate and an error message will be generated.
• SEAM3D assumes evaporation is insignificant in cells and for stress periods where
phytoremediation is active. If this is not true, model results will be less accurate.
• The location and flow rate of discharge is obtained from the flow model directly through the
unformatted flow-transport link file.
207
Table 8.1. Transpiration Stream Concentration Factors (TSCF) and Root Concentration Factors (RCF) for selected ground-water contaminants.
Log Kow TSCF RCF (L/kg)
Benzene, (C6H6) 2.13 0.71 3.6 Toluene, (C7H8) 2.65 0.74 4.5
Ethylbenzene, (C8H10) 3.13 0.63 6.0 m-Xylene, (C6H4(CH3)2) 3.20 0.61 6.2
o-Xylene, (C6H4(CH3)2) 2.95 0.70 4.9 MTBE, (C5H12O) 1.20 0.41 3.2 1,3,5-TMB, (C9H12) 3.42 0.56 6.5 PCE, (C2Cl4) 3.14[3] 5.96 0.37 TCE, (C2HCl3) 2.33[1] 0.75[2] 3.0[2]
1,2-cisDCE, (C2H2Cl2) 1.86[3] 3.44 0.78 Vinyl Chloride, (C2H3Cl) 1.23[3] 3.17 0.69 Perchlorate (ClO4) -7.18[4] 3.00 - TCA, (C2H3Cl3) 2.49[3] 4.12 0.64 Tetrachloromethane, CTC (CCl4) 2.6[5] 4.32 0.60 Naphthalene, (C10H8) 3.37 0.56 7.2 Acenaphthylene, (C12H8) 4.33 0.21 20.6 Acenaphthene, (C12H10) 4.07 0.29 14.9 Fluorene, (C13H10) 4.18 0.25 17.0 Phenanthrene, (C14H10) 4.46 0.17 24.3
Anthracene, (C14H10) 4.45 0.17 24.0 1 Physical chemical properties (Schwarzenbach, et al., 1993) 2 Measured data from hydroponic studies with hybrid poplars (Burken and Schnoor, 1998; Dietz and Schnoor, 2001). 3 Arthur D. Little, Inc. (1987). The installation restoration program toxicology guide, Volume 1. Section 2:1-16. 4 ITRC 2002, http://www.itrcweb.org/user/isb-8r.pdf 5 The International Uniform Chemical Information Database (IUCLID), 1996
208
Example *.pup input files
Freundlich, ISO=2
1 T T 2 … Root sorption is OFF (F), direct uptake is ON (T), and ISOTHMP=1
2 0 1750000. … Array for the root bulk density. Only if FRCF = T 3 0 0.80 … Array for the Freundlich exponent N, if Isotherm=2 (> 1) for
species 1 0 1.00 … Array for the Freundlich exponent N, if Isotherm=2 (>1) for
species 2 4 … …
5 1 1 1 … INSURF, INEXDP, INRCF (reading flags for surface elevation, extinction depth, and RCF)
6 0 8.0 … Surface Elevation array 7 0 4.0 … Extinction depth array 8 0 0.0 … RCF value for species #1 0 0.0 … RCF value for species #1 9 1 … INTSCF is the TSCF read flag. 10 0 1.00 … Array for TSCF value (=1.0) stress period #1, species 1 0 0.00 … Array for TSCF for species 2
Langmuir, ISO=3
1 T T 3 … Root sorption is OFF (F), direct uptake is ON (T), and ISOTHMP=1
2 0 1750000. … Array for the root bulk density. Only if FRCF = T 3 0 0.80 … Array for the Langmuir equilibrium constant (Kl), if
Isotherm=3 for species 1 0 1.0 … Array for the Langmuir equilibrium constant (Kl), if
Isotherm=3 for species 2 4 0 8.0 … Array for the total concentration of the plant uptake sites
available, if Isotherm=3 (Tc)for species #1
0 1.0 … Array for the total concentration of the plant uptake sites available, if Isotherm=3 (Tc)for species #2
5 1 1 1 … INSURF, INEXDP, INRCF (reading flags for surface elev, extinction depth, and RCF)
6 0 8.0 … Surface Elevation array 7 0 4.0 … Extinction depth array 0 0.0 … RCF value for species #1 0 0.0 … RCF value for species #1 9 1 … INTSCF is the TSCF read flag. 10 0 1.00 … Array for TSCF value (always=1.0 in case of ISO=3) for stress
period #1, species 1 0 0.00 … Array for TSCF for stress period #1, species 2
209
BIBLIOGRAPHY
Aitchison, E. W., S. L. Kelley, et al. (2000). "Phytoremediation of 1,4-dioxane by hybrid poplar trees." Water Environment Research.
Allen, R. G., L. S. Pereira, et al. (1998). "Crop evapotranspiration - Guidelines for computing crop
water requirements." FAO Irrigation and drainage paper 56. Anderson, M. P. and W. W. Woessner (1992). Applied Groundwater Modeling: Simulation of Flow
and Advective Transport, Academic Press, INC. Anderson, T. A., E. A. Guthrie, et al. (1993). "Bioremediation in the rhizosphere: plant roots and
associated microbes clean contaminated soil." Environmental Science & Technology 27: 2630-2636.
ASCE, A. S. o. C. E. (1990). Evapotranspiration and Irrigation Water Requirements. Manuals and
Reports on Engineering Practice. No. 70. Baker, A. J. M. and R. R. Brooks (1989). "Terrestrial higher plants which hyperaccumulate metal
elements - a review of their distribution, ecology and phytochemistry." Biorecovery 1: 81-126. Bedient, P. H., H. S. Rifai, et al. (1994). Ground Water Contamination: Transport and Remediation.
NJ, Prentice Hall, Englewood Cliffs. Behrendt, H., R. Bruggemann, et al. (1995). "Numerical and Analytical Model of Pesticide Root
Uptake: Model Comparison and Sensitivities." Chemosphere 30(10): 1905-1920. Berti, W. R. and S. D. Cunningham (2000). Phytostabilization of metals. Phytoremediation of toxic
metals - using plants to clean-up the environment. NY, John Wiley & Sons, Inc. Black, H. (1995). "Absorbing possibilities: phytoremediation." Environmental Health Perspectives
103(12): 1106-1108. Blaylock, M. J., D. E. Salt, et al. (1997). "Enhanced accumulation of Pb in Indian Mustard by soil-
applied chelating agents." Environmental Science & Technology 31(3): 860-865. Boersma, L., F. T. Lindstrom, et al. (1990). "Model for uptake of organic chemicals by plants." Oregon
State University, Agricultural Experiment Station Bulletin. Briggs, G. G., R. H. Bromilow, et al. (1982). "Relationships Between Lipophilicity and Root Uptake
and Translocation of Non-ionised Chemicals by Barley." Pesticide Science 13: 495-504.
210
Briggs, G. G., R. H. Bromilow, et al. (1983). "Relationships Between Lipophilicity and the Distribution of Non-ionised Chemicals in Barley Shoots Following Uptake by the Roots." Pesticide Science 14: 492-500.
Brikowski, T. (2002). Quantification and Measurement of the Hydrologic Cycle. GEOS 4310/5310
Lecture Notes. Burbey, T. J. (2003). Numerical Modeling of Groundwater Flow and Transport. Lecture notes.
Blacksburg, VA. Burken, J. G. and J. L. Schnoor (1997). "Uptake and metabolism of atrazine by poplar trees."
Environmental Science & Technology 31: 1399-1406. Burken, J. G. and J. L. Schnoor (1998). "Predictive relationships for uptake of organic contaminants by
hybrid poplar trees." Environmental Science & Technology 32: 3379-3385. Calamari, D., M. Vighi, et al. (1987). "The use of terrestrial plant biomass as a parameter in the fugacity
model." Chemosphere 16: 2359-2364. Cammarota, A. V., Jr. (1980). Production and uses of zinc. Zinc in the environment. Part I: Ecological
Cycling. New York, John Wiley & Sons. Carl, E. G., B. Faulkner, et al. (1999). In Situ Enhanced Source Removal, EPA. Castro, S., L. C. Davis, et al. (2000). Phytoremediation of aircraft deicer and antifreeze formulations.
Conference on hazardous Waste Research. Castro, S., L. C. Davis, et al. (2001). "Plant-enhanced remediation of glycol-based aircraft deicing
fluids." Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 5(3): 141-152. CBS, N. (2003). Water facts and figures, International statistics. Chaney, R. L. (1983). Plant uptake of inorganic waste constitutes. Land treatment of hazardous wastes.
P. B. M. J.F. Parr, and J.M. Kla. Park Ridge, NJ, Noyes Data Corp. Chappell, J. (1997). Phytoremediation of TCE in Groundwater using Populus, U.S. EPA Technology
Innovation Office. Charbeneau, R. J. (2000). Groundwater Hydraulics and Pollutant Transport. New Jersey, Prentice Hall. Chiou, C. T. (2002). Partition and Adsorption of Organic Contaminants in Environmental Systems.
New Jersey, John Wiley & Sons, Inc. Chiou, C. T., G. Sheng, et al. (2001). "A partition-limited model for the plant uptake of organic
contaminants from soil and water." Environmental Science Technology 35: 1437–1444.
211
Compton, H. R., D. M. Haroski, et al. (1998). Pilot-scale use of trees to address VOC contamination. Bioremediation and Phytoremediation: Chlorinated and Recalcitrant Compounds. Columbus, OH, Battelle Press: 245-250.
Cook, J. (1977). "Environmental pollution by heavy metals." International Journal of Environmental
Studies 10(4): 253-266. Cunningham, S. and W. R. Berti (1993). "Remediation of contaminated soils with green plants: An
Overview." In Vitro Cellular & Developmental biology Plant 29P(4): 227 -2 32. Cunningham, S. D., W. R. Berti, et al. (1995). "Phytoremediation of contaminated soils." Trends in
Biotechnology 13(9): 393-397(5). Cunningham, S. D., J.R. Shann, D.E. Crowley, and T.A. Anderson (1997). Phytoremediation of
contaminated water and soil. Phytoremediation of soil and water contaminants, Orlando, Florida American Chemical Society Publication
Cunningham, S. D. and D. W. Ow. (1996). "Promises and prospects of phytoremediation." Plant
physiology 110: 715-719. Curl, E. A. and B. Truelove (1986). The rhizosphere. New York, Springer-Verlag. Danskin, W. R. (1988). Preliminary evaluation of the hydrogeologic system in Owens Valley, California.
USGS, Water-Resources Investigation Report WRI 88-4003. Darcy, H. (1856). "Les Fountaines de la Ville de Dijon." Davis, L. C., M. K. Banks, et al. (1998). Plant-Based Bioremediation. Bioremediation: Principles and
Practice. S. K. S. a. R. L. Irvine. Lancaster, PA, Technomic Publ. Co. II: 183-219. Davis, L. C., L. E. Erickson, et al. (1993). "Effects of Plants on the Bioremediation of Contaminated
Soil and Ground Water." Environmental Progress 12(1): 67-75. Dennis, R. M., D. Dworkin, et al. (1994). Soil-washing processes for site remediation. Remediation of
hazardous waste contaminated soils. D. L. W. a. D. J. Trantolo. New York, Marcel Dekker, Inc. DeSouza, M. P., E. A. H. Pilon-Smits, et al. (2000). The physiology and biochemistry of selenium
volatilization by plants. Phytoremediation of toxic metals using plants to clean-up the environment. I. R. a. B. D. Ensley. New York, John Wiley & Sons, Inc.
Dietz, A. C. (2000). Phytotoxicity of chlorinated aliphatics to hybrid poplar cuttings and axenic tissue
cultures. Civil & Environmental Eng., University of Iowa. Ph. D.: 166. Dietz, A. C. and J. L. Schnoor (2001). "Advances in Phytoremediation." Environmental Health
Perspectives 109(S1): 163-168.
212
Dushenkov, S. and Y. Kapulnik (2000). Phytofilitration of metals. Phytoremediation of toxic metals - using plants to clean-up the environment. I. R. a. B. D. Ensley. New York, John Wiley & Sons, Inc.
Dushenkov, V., P. B. A. N. Kumar, et al. (1995). "Rhizofiltration: the use of plants to remove heavy
metals from aqueous streams." Environmental Science & Technology 29(5): 1239-1245. Ensen, M. E., R. D. Burman, et al. (1990). Evapotranspiration and Irrigation Water Requirements.
New York, ASCE Manuals and Reports of Engineering Practice No. 70. Ensley, B. D. (2000). Rational for use of phytoremediation. Phytoremediation of toxic metals - using
plants to clean-up the environment. I. R. a. B. D. Ensley. New York, John Wiley & Sons, Inc. EPA (1998). A Citizen's Guide to Phytoremediation, EPA. EPA (1999). Phytoremediation resource guide, U.S. Environmental Protection Agency: 56. EPA (2001). Treatment Technologies for Site Cleanup: Annual Status Report, EPA. Erickson, L. E. M. K. B. L. C. D. A. P. S. M. N. K. R. a. J. C. T. (1994). "Using Vegetation to Enhance
In Situ Bioremediation." Environmental Progress 13: 226-234. Fayer, M. J. (2002). UNSAT-H Version 3.0: Unsaturated Soil Water and Heat Flow Model, Theory,
User Manual, and Examples, United States Department Of Energy. Feddes, R. A. (1981). "Water use models for assessing root zone modification. Modifying the Plant
Root Environment." Monograph 4, ASCE: 347–390. Feddes, R. A., Hoff, H., Bruen, M., Dawson, T.E., de Rosnay P., Dirmeyer, Jackson, R.B., Kabat, P.,
Kleidon, A., Lilly, A ., and Pitman, A.J. (2001). "Modeling root water uptake in hydrological and climate models." Bulletin of the American Meteorological Society 82: 2797-2810.
Feddes, R. A., S. P. Neuman, et al. (1975). "Finite Element Analysis of Two-dimensional Flow in Soils
Considering Water Uptake by Roots: II. Field Applications." Soil Science Society of America Journal 39: 231-237.
Ferro, A. M., R. C. Sims, et al. (1994). "Hycrest Crested Wheatgrass Accelerates the Degradation of
Pentachlorophenol in Soil." Journal of Environmental Quality 23(2): 272-279. Fetter, C. W. (1994). Applied Hydrogeology, 3rd ed. NJ, Prentice Hall Inc. Fetter, C. W. (1999). Contaminant Hydrogeology. NJ, Prentice Hall Inc. Fetterolf, G. J. I. (1998). Characterization of a Creosote-Contaminated Tie Yard Site and the Effects of
Phytoremediation. CEE. Blacksburg, Virginia Tech. M. Sc.
213
Focazio, M. J., T. E. Reilly, et al. (2003). Assessing Ground-Water Vulnerability to Contamination: Providing Scientifically Defensible Information for Decision Makers. U.S. Geological Survey Circular 1224, U.S. Geological Survey: 30.
Freyer, M. E. and C. D. Collins (2003). "Model Intercomparison for the Uptake of Organic Chemicals
by Plants." Environmental Science & Technology 37: 617-1624. Gallagher, D. (1998). "Groundwater Pollution Premier." Retrieved 2/18/05, 2005, from http://ewr.cee.vt.edu/environmental/teach/gwprimer/gwprimer.html. Glass, D. J. (2000). Economic potential of phytoremediation. Phytoremediation of toxic metals - using
plants to clean-up the environment. I. R. a. B. D. Ensley. New York, John Wiley & Sons, Inc. Hanson, R. L. (1991). "Evapotranspiration and Droughts, in Paulson." National Water Summary 1988-
89--Hydrologic Events and Floods and Droughts: 99-104. Harris, C. R. and W. W. Sans (1967). "Absorption of organochlorine insecticide residues from
agricultural soils by root crops." Journal of agricultural and food chemistry 15: 861–863. Heaton , A. C. P., C. L. Rugh, et al. (1998). "Phytoremediation of mercury- and methylmercury-
polluted soils using genetically engineered plants." Journal of Soil Contamination, 7(4): 497-509. Henry, J. R. (2000). An Overview of the Phytoremediation of Lead and Mercury, EPA, Office of Solid
Waste and Emergency Response, Technology Innovation office. Hinchman, R. R. (1996). Phytoremediation: Using Green Plants To Clean Up Contaminated Soil,
Groundwater, And Wastewater. International Topical Meeting on Nuclear and Hazardous Waste Management, Seattle WA.
Homas, J. M., S. M. Carlton, et al. (1989). Ground-water hydrology and simulated effects of
development in Smith Creek Valley, a hydrologically closed basin in Lander County, Nevada. U.S. Geological Survey Professional Paper, U.S. Geological Survey: 57.
Hong, M. S., W. F. Farmayan, et al. (2001). "Phytoremediation of MTBE from a groundwater plume."
Environmental Science & Technology 35(6): 1231-1239. Hughes, J. B., J. Shanks, et al. (1997). "Transformation of TNT by aquatic plants and plant tissue
cultures." Environmental Science & Technology 31: 266-271. International Joint Commission (1970). Pollution of Lake Erie, Lake Ontario, and the international
section of the St. Lawrence River. ITRC (1999). Phytoremediation Decision Tree, Interstate Technology and Regulatory Cooperation
Work Group, Phytoremediation Work Team.
214
ITRC (2001). Phytotechnology Technical and Regulatory- Guidance Document, Technical/Regulatory Guidelines, Interstate Technology and Regulatory Cooperation Work Group, Phytotechnologies Work Team.
ITRC (2002). A systematic Approach to In situ Bioremediation In Groundwater, Interstate
Technology and Regulatory Cooperation Work Group, Phytotechnologies Work Team. Jensen, M. E. (1973). "Consumptive Use of Water and Irrigation Water Requirements." American
Society of Civil Engineering. Jordahl, J. L., L. Foster, et al. (1997). "Effect of hybrid poplar trees on microbial populations important
to hazardous waste bioremediation." Environmental Toxicology and Chemistry 16: 1318-1321. Keller, A. A. (2003). "The Workshop on Tools for Environmental Cleanup: Engineered Plants for
Phytoremediation." International journal of phytoremediation. Kijune, S., M. Y. Corapcioglu, et al. (2001). "Plant and Environment Interactions: Plant Contamination
by Organic Pollutants in Phytoremediation." Journal of Environmental Quality 30: 2081-2090. Klocke, N. L., K. G. Hubbard, et al. (1996). "Evapotranspiration (ET) or Crop Water Use." from
http://www.p2pays.org/ref/20/19769.htm. Kroes, J.G., J.C. van Dam, J. Huygen and R.W. Vervoort, 1999. User’s Guide of SWAP version 2.0;
Simulation of water flow, solute transport and plant growth in the Soil-Water-Atmosphere-Plant environment. Wageningen Agricultural University. Report 81, DLO Winand Staring Centre. Technical Document 53.128 pp.
Kumar, P. B. A. N., V. Dushenkov, et al. (1995). "Phytoextraction: The use of plants to remove heavy
metals from soils." Environmental Science & Technology 29(1232-1238). Landmeyer, J. E. (2001). "Monitoring the Effect of Poplar Trees on Petroleum-Hydrocarbon and
Chlorinated-Solvent Contaminated Groundwater." International Journal of Phytoremediation 3: 61-85.
Lathman, P. E. and G. R. Lanza (1998). "Phytoremediation: current views on an emerging green
technology." Journal of Soil Contamination 7(4): 415-432. Lay, D. J. (1999). "Phytoremediation of Trichloroethylene (TCE)." Retrieved 9/17/2006, 2006, from
http://horticulture.coafes.umn.edu/vd/h5015/99fpapers/lay.htm. Lee, D. B. (1989). Tragedy in Alaska Waters. National Geographic. 176: 260-264. Lee, M. D., J. M. Thomas, et al. (1988). "Biorestoration of Aquifers Contaminated with Organic
Compounds." CRC Critical Reviews in Environmental Control 18(1): 29-89.
215
Leo, A., C. Hansch, et al. (1971). "Partition Coefficients and Their Uses." Chemical Reviews 71(6): 525-621.
Levine, A. G. (1982). Love Canal: science, politics, and people. Lexington, MA, Lexington Books. Lichtenstein, E. P. (1959). "Absorption of some chlorinated hydrocarbon insecticides from soils into
various crops." Journal of Agriculture and Food Chemistry 7: 430–433. Long, G. M. (1993). "Clean Up Hydrocarbon Contamination Effectively." Chemical Engineering
Progress 89(5). Ma, X. and J. Burken (2002). "VOCs Fate and Partitioning in Vegetation: Use of Tree Cores in
Groundwater Analysis." Environmental Science & Technology 36(21): 4663 -4668. Mariño, M. A. and J. C. Tracy (1988). "Flow of water through root-soil environment." Journal of
Irrigation and Drainage Engineering, ASCE 114(4): 588-604. Matso, K. (1995). "Mother Nature’s Pump and Treat." Civil Engineering, ASCE 65(10): 46-49 Matthews, D. W., J. Massmann, et al. (2002). "Influence of aquifer properties on Phytoremediation
Effectiveness." Groundwater 41(1): 41-47. McDonald, M. G. and A. W. Harbaugh (1988). A modular three-dimensional finite-difference ground-
water flow model. Techniques of Water-Resources Investigations, U.S. Geological Survey. Medina, V. F., E. Maestri, et al. (2003). Plant Tolerances to Contaminants. Phytoremediation:
Transformation and control of contaminants. J. L. Schnoor and A. Zehnder. New Jersey, John Wiley & Sons, Inc.: 1024.
Miller, R. R. (1996). Phytoremediation: Technology Overview Report. GWRTAC, Ground Water
Remediation Technologies Analysis Center: 26. Molz, F. J. (1981). "Models of water transport in the soil–plant system: A review." Water Resources
Research 17(5): 1245-1260. Muller, J. F., D. W. Hawker, et al. (1994). "Calculation of Bioconcentration Factors of Persistent
Hydrophobic Compounds in the Air/Vegetation system." Chemosphere 29: 623-640. Narayanan, M., C. D. Davis, et al. (1995). "Experimental and modeling studies of the fate of organic
contaminants in the presence of alfalfa plants." Journal of Hazardous Materials 41: 229-249. Newman, L. A., S. E. Strand, et al. (1997). "Uptake and Biotransformation of Trichloroethylene by
Hybrid Poplars." Environmental Science & Technology 31(4): 1062-1067. Newman, L. A., X. Wang, et al. (1999). "Remediation of trichloroethylene in an artificial aquifer with
trees: a controlled case study." Environmental Science & Technology 33: 2257-2265.
216
NRC (1999). Metals and radionuclides: technologies for characterization, remediation, and
containment. Groundwater & soil clean up: improving management of persistent contaminants. Washington, D.C., National Academy Press.
NRC, C. o. I. R. T. (1997). Challenges of groundwater and soil cleanup. Innovations in Ground Water
and Soil Cleanup: From Concept to Commercialization. Washington, D.C., National Academies Press: 292.
Nyer, E. K. and E. G. Gatliff (1996). "Phytoremediation " Ground Water Monitoring and Remediation
16(1): 58-62. Paterson, S. and D. Mackay (1994). "A Model of Organic Chemical Uptake by Plants from Soil and the
Atmosphere." Environmental Science & Technology 28: 2259-2266. Pollock, D. W. (1994). User's guide for MODPATH/MODPATH-PLOT Version 3: A particle
tracking post-processing package for MODFLOW, the U.S. Geological Survey finite-difference ground-water flow model, U.S. Geological Survey.
Quinn, J. J. (2000). Modeling Plume Capture at Argonne National Laboratory – East. Argonne, IL
60439, Environmental Assessment Division, Argonne National Laboratory. Raskin, I. and B. D. Ensley, Eds. (2000). Phytoremediation of toxic metals - using plants to clean-up
the environment. NY, John Wiley & Sons, Inc. Reed, D. T., I. R. Tasker, et al. (1992). Environmental restoration and separation science.
Environmental Remediation: Removing Organic and Metal Ion Pollutants (ACS Symposium S.) D. T. R. G. F. Vandegrift, and I. R. Tasker. Washington, DC., American Chemical Society ACS Symposium Series 509.
Reeves, R. D. and A. J. M. Baker (2000). Metal-accumulating plants. Phytoremediation of toxic metals -
using plants to clean-up the environment. I. R. a. B. D. Ensley. New York, John Wiley & Sons, Inc.: 193-230.
Rugh, C. L., S. P. Bizily, et al. (2000). Phytoreduction of environmental mercury pollution.
Phytoremediation of toxic metals - using plants to clean-up the environment. I. R. a. B. D. Ensley. New York, John Wiley & Sons, Inc.
Salt, D. E., M. Blaylock, et al. (1995). "Phytoremediation: a novel strategy for the removal of toxic
metals from the environment using plants." Biotechnology 13: 468-474. Schnoor, J. L. (2002). Phytoremediation of Soil and Groundwater. GWRTAC Ground-Water
Remediation Technologies Analysis Center. Schnoor, J. L., L. A. Light, et al. (1995). "Phytoremediation of organic and nutrient contaminants."
Environmental Science & Technology 29(7): 318A-323A.
217
Schnoor, J. R. (1997). Phytoremediation: Evaluation report. GWRTAC Pittsburgh, Pennsylvania,
Ground-Water Remediation Technologies Analysis Center. Shimp, J. F., J. C. Tracy, et al. (1993). "Beneficial effects of plants in the remediation of contaminated
soil and ground water." Critical Reviews in Environmental Control 23(1): 41-47. Shuttleworth, W. J. (1993). Evaporation. Handbook of Hydrology. D. R. Maidment, McGraw-Hill. Simbnek, J., K. Huang, et al. (1995). SWMS_3D, 3D Code for Simulating Water Flow and Solute
Transport in Three-Dimensional Variably-Saturated Media. Riverside, California, U. S. Salinity laboratory, Agricultural research service, U. S. Department of agriculture.
Stomp, A. M., K. H. Han, et al. (1994). "Genetic Strategies for Enhancing Phytoremediation." Annals
of the New York Academy of Science 721: 481-491. Struckhoff, C. G. and G. J. Burken (2005). "Vapor-Phase Exchange of Perchloroethene between Soil
and Plants " Environmental Science & Technology 39(6): 1563 -1568. Suszcynsky, E. M. and J. R. Shann (1995). "Phytotoxicity and accumulation of mercury subjected to
different exposure routes." Environmental Toxicology and Chemistry 14: 61-67. Taiz, L. and E. Zeiger (2002). Plant Physiology. Sunderland, MA, Sinauer Associates. Thoma, G. J., T. B. Lam, et al. (2003). "A Mathematical Model of Phytoremediation for Petroleum-
Contaminated Soil: Model Development." International Journal of Phytoremediation 5(1): 41-55. Thompson, P. L., L. A. Ramer, et al. (1998). "Uptake and transformation of TNT by hybrid poplar
trees." Environmental Science & Technology 32: 975-980. Thornthwaite, C. W. (1944). "Report of the committee on transpiration and evaporation."
Transactions, American Geophysical Union 25: 687. Topp, E., L. Scheunert, et al. (1986). "Factors affecting the uptake of 14C-labeled organic chemicals by
plants from soil." Ecotoxicology and environmental safety 11(2): 219-228. Trapp, S. (1995). Model for Uptake of Xenobiotics into Plants. Plant Contamination: Modeling and
Simulation of Organic Chemical Processes. S. Trapp, McFarlane, J. C. Boca Raton, FL, Lewis Publishers: 107-151.
Trapp, S. (2000). "Modeling uptake into roots and subsequent translocation of neutral and ionisable
organic compounds." Pest Management Science 56: 767-778. Trapp, S. (2004). "Plant Uptake and Transport Models for Neutral and Ionic Chemicals." ESPR –
Environmental science and pollution research international 11(1): 33-39.
218
Trapp, S. (2004). Uptake of Neutral Organic Chemicals into Plants Models, equations and exercises. PACE PhD course on ecotoxicology Sardinia 5 to 9 July 2004: 30.
Trapp, S. and M. Matthies (1995). "Generic One-Compartment Model for Uptake of Organic
Chemicals by Foliar Vegetation." Environmental Science & Technology 29: 2333-2338. Trapp, S. and M. Matthies (1998). Chemodynamics and Environmental Modelling: An Introduction.
Berlin ; New York Springer. Trapp, S., J. C. McFarlane, et al. (1995). Model for Uptake of Xenobiotics into Plants. Plant
Contamination: Modeling and Simulation of Organic Chemical Processes. S. Trapp, McFarlane, J.C., Lewis Publishers: 107-151.
Trapp, S., McFarlane, J. C., Matthies, M. (1995). "Model for Uptake of Xenobiotics into Plants:
Validation with Bromacil Experiments." Environmental Toxicology and Chemistry 13(3): 413-422. Travis, C. C. and A. D. Arms (1988). "Bioconcentration of Organics in Beef, Milk and Vegetation."
Environmental Science & Technology 22: 271-274. Unland, H. E., A. M. Arain, et al. (1998). "Evaporation from a Riparian System in a Semi-Arid
Environment." Hydrological Processes 12: 527-542. USBR U. S. Bureau of Reclamation (1991). Inventory of Hydrologic Models. Global Climate Change
Response Program. Denver, Colorado United States Department of the Interior. USDA (1971). Irrigation Water Requirement, SCS Technical Release No. 21, U.S. Department of
Agriculture Soil Conservation Service. USGS. (2003). "Phytoremediation definition page." Retrieved 9/18/06, from http://toxics.usgs.gov/definitions/phytoremediation.html. USGS, a. (1996). "Earth's water distribution." Retrieved 9/18/06, from http://ga.water.usgs.gov/edu/waterdistribution.html. Vroblesky, D. A., C. T. Nietch, et al. (1999). "Chlorinated Ethenes from Groundwater in Tree Trunks."
Environmental Science & Technology 33(3): 510 -515. Waddill, D. W. and M. A. Widdowson (1998). SEAM3D, A Numerical Model for Three-Dimensional
Solute Transport Coupled to Sequential Electron Acceptor-Based Biological Reactions in Groundwater, US Army Corps of Engineers.
Walton, B. T. and T. A. Anderson (1990). "Microbial Degradation of Trichloroethylene in the
Rhizosphere: Potential Application to Biological Remediation of Waste Sites." Applied and Environmental Microbiology 56(4): 1012-1016.
219
Ward, C. H. (1999). Preface In Groundwater and soil cleanup: improving management of persistent contaminants. Washington, DC., National Academy Press.
Watanabe, M. E. (1997). "Phytoremediation on the brink of commercialization." Environmental.
Science & Technology 31(4): 182A-186A. Widdowson, M. A., D. W. Waddill, et al. (2002). SEAM3D v3: A Numerical Model for Three-
Dimensional Solute Transport Coupled to Sequential Electron Acceptor-Based Biological Reactions in Groundwater., Virginia Polytechnic Institute and State University.
Wood, A. (2003). Remediation Control Strategies and Cost Data for an Economic Analysis of a
Mercury Total Maximum Daily Load in California. U.S.G.S Open-File Report, USGS. Wood, P. A. (1997). Remediation methods for contaminated sites. Issues in environmental science and
technology: contaminated land and its reclamation. R. E. H. a. R. M. Harrison. Letchworth, U.K., The Royal Society of Chemistry.
WRRC. (2004). "phreatophyte." Retrieved 9/18/06, from http://www.ag.arizona.edu/AZWATER/reference/glossary/phreat.html. Wullschleger, S. D., F. C. Meinzer, et al. (1998). "A review of whole-plant water use studies in trees."
Tree Physiology 18(8): 499-512. Ying, O. (2002). "Phytoremediation: modeling plant uptake and contaminant transport in the soil-
plant-atmosphere continuum." Journal of Hydrology 266: 66-82. Zheng, C. and D. G. Bennett (2002). Applied Contaminant Transport Modeling, 2nd Edition. New
York, John Wiley & Sons. Zheng, C. and P. P. Wang (1998). MT3DMS: A Modular Multi-Species Three-Dimensional Transport
Model, Documentation and User’s Guide, U.S. Army Corps of Engineers Waterway Experiment Station.
220
Appendix A
Auxiliary Figures and Tables from Chapter 5
W(ET)=300, L(ET)=Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, g
ET, W=300TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
LET=Lp, QET = 0.0005 m3/d/m2 a) W=300
W(ET)=250, L(ET)=Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
GMS-ETTSCF=1.00TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
W(ET)=200, L(ET)=Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
GMS-ETTSCF=1.00TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
b) W=250 c) W=200
W(ET)=150, L(ET)=Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
ETTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.00
W(ET)=100, L(ET)=Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
GMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
d) W=150 e) W=100
Figure A.1. Effect of ET width on solute mass removal, LET=Lp.
221
W(ET)=300, L(ET)=0.5Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, Days
Mas
s-in
, g
ET, W=300TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
LET=0.5Lp, QET = 0.001 m3/d/m2 a) W=300
W(ET)=250, L(ET)=0.5Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
GMS-ETTSCF=1.00TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
W(ET)=200, L(ET)=0.5Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
GMS-ETTSCF=1.00TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
b) W=250 c) W=200
W(ET)=150, L(ET)=0.5Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
ETTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.00
W(ET)=100, L(ET)=0.5Lp
20
22
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
GMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
d) W=150 e) W=100
Figure A.2. Effect of ET width on solute mass removal, LET=0.5Lp.
222
L(ET)=Lp, TSCF=1.0
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
LET=Lp, QET = 0.0005 m3/d/m2 TSCF=1.0
L(ET)=Lp, TSCF=0.75
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
L(ET)=Lp, TSCF=0.50
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
TSCF=0.75 TSCF=0.50
L(ET)=Lp, TSCF=0.25
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
L(ET)=Lp, TSCF=0.0
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
TSCF=0.25 TSCF=0.0
Figure A.3. Effect of TSCF on solute mass removal, LET=Lp.
223
L(ET)=0.5Lp, TSCF=1.0
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
LET=0.5Lp, QET = 0.001 m3/d/m2 TSCF=1.0
L(ET)=0.5Lp, TSCF=0.75
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
L(ET)=0.5Lp, TSCF=0.50
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
TSCF=0.75 TSCF=0.50
L(ET)=0.5Lp, TSCF=0.25
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
L(ET)=0.5Lp, TSCF=0.0
24
26
28
30
32
34
36
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s-in
, g
W=300W=250W=200W=150W=100NA
TSCF=0.25 TSCF=0.0
Figure A.4. Effect of TSCF on solute mass removal with different ET lengths.
224
T=10 yr, L(ET)=0.5Lp
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L TSCF=1.0
TSCF=0.75TSCF=0.5TSCF=0.0NA
T=5 yr, L(ET)=0.5Lp
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L TSCF=1.0
TSCF=0.75TSCF=0.5TSCF=0.0NA
T=10 yr, L(ET)=Lp
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L TSCF=1.0
TSCF=0.75TSCF=0.5TSCF=0.0NA
T=5 yr, L(ET)=Lp
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L TSCF=1.0
TSCF=0.75TSCF=0.5TSCF=0.0NA
Figure A.5. Concentration profiles along the length of the plume for different values of TSCF
at different simulation times (5 yr, and 10 yr).
W=300, L(ET)=Lp
0.0001
0.001
0.01
0.1
1
10
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+365t=+1825t=+3650
W=300, L(ET)=0.5Lp
0.0001
0.001
0.01
0.1
1
10
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
t=+365t=+1825t=+3650
Figure A.6. Concentration vs. distance at different observation points downstream the source
(with exponential fitting in the bottom charts).
225
W/Ws=3.0, L(ET)=Lp
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
LET=Lp, QET = 0.0005 m3/d/m2 (a)
W=250, L(ET)=Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
W=200, L(ET)=Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
75
(b) (c)
W=150, L(ET)=Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
W=100, L(ET)=Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
(d) (e)
Figure A.7. Concentration profiles for different TSCF values used to calculate the plume
length at a concentration = 1% of the source concentration for LET=Lp.
226
W/Ws=3.0, L(ET)=0.5Lp
0.0001
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
LET=0.5Lp, QET = 0.001 m3/d/m2 (a)
W=250, L(ET)=0.5Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
W=200, L(ET)=0.5Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
75
(b) (c)
W=150, L(ET)=0.5Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
W=100, L(ET)=0.5Lp
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAGMS-ETTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
(d) (e)
Figure A.8. Concentration profiles for different TSCF values used to calculate the plume
length at a concentration = 1% of the source concentration for LET=0.5Lp.
227
W=300, L(ET)=Lp
-5
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Thou
sand
s
Dist., m
Av.,
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
a) LET=Lp, QET = 0.0005 m3/d/m2 WET=300
W=250, L(ET)=Lp
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
W=200, L(ET)=Lp
0
5000
10000
15000
20000
25000
30000
35000
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
WET=250 WET=200
W=150, L(ET)=Lp
0
5000
10000
15000
20000
25000
30000
35000
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist.,m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
W=100, L(ET)=Lp
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000
1100
1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
WET=150 WET=100
Figure A.9 Average Mass-flux results at different cross-sections downstream the source for
LET=Lp and different values of WET and TSCF.
228
W=300, L(ET)=0.5Lp
-5
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100
Thou
sand
s
Dist., m
Av.,
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
LET=0.5Lp, QET = 0.001 m3/d/m2 WET=300
W=250, L(ET)=0.5Lp
0
5000
10000
15000
20000
25000
30000
35000
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
W=200, L(ET)=0.5Lp
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/L NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
WET=250 WET=200
W=150, L(ET)=0.5Lp
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.00
W=100, L(ET)=0.5Lp
0
5
10
15
20
25
30
35
0 100 200 300 400 500 600 700 800 900 1000 1100
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d NA
TSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
WET=150 WET=100
Figure A.10. Average Mass-flux results at different cross-sections downstream the source for
LET=0.5Lp and different values of WET and TSCF.
229
26000
27000
28000
29000
30000
31000
32000
33000
34000
35000
36000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Mas
s re
mov
al, g
NAGMSTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
33000
33200
33400
33600
33800
34000
34200
34400
34600
34800
35000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Mas
s re
mov
al, g
NAGMSTSCF=1.0TSCF=0.75TSCF=0.5TSCF=0.25TSCF=0.0
ET at the left edge ET at the right edge
TSCF=0.50
26
27
28
29
30
31
32
33
34
35
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s re
mov
al, g
LEFT EdgeRIGHT Edge
TSCF=0.75
26
27
28
29
30
31
32
33
34
35
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
Time, d
Mas
s re
mov
al, g
LEFT EdgeRIGHT Edge
TSCF=0.5 TSCF=0.75
Figure A.11. Effect of the phytoremediation location and TSCF on solute mass removal.
230
Q=150
0.0001
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L L(ET)=0.25 Lp
L(ET)=0.50 LpL(ET)=0.75 LpL(ET) = LpNA
(a) ET starts at the left edge
Q=150
0.0001
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L L(ET) = 0.25 Lp
L(ET) = 0.50 LpL(ET) = 0.75 LpL(ET) = LpNA
(b) ET starts at the right edge
Figure A.12. Concentration profiles for different aquifer in-flux (Qin=1.50 m3/d/cell) and ET
lengths
231
Q=105
0.00001
0.0001
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L L(ET)=0.25 Lp
L(ET)=0.50 LpL(ET)=0.75 LpL(ET) = LpNA
(a) ET starts at the left edge
Q=105
0.00001
0.0001
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L L(ET)=0.25 Lp
L(ET)=0.50 LpL(ET)=0.75 LpL(ET) = LpNA
(b) ET starts at the right edge
Figure A.13. Concentration profiles for different aquifer in-flux (Qin=1.05 m3/d/cell) and ET
lengths
232
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L
NAGMSTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
0.001
0.01
0.1
1
0 200 400 600 800 1000 1200
Dist., m
Conc
., m
g/L
NAGMSTSCF=1.0TSCF=0.75TSCF=0.50TSCF=0.25TSCF=0.0
ET at the left edge ET at the right edge (a)
L(ET)=0.5Lp, TSCF=0.50
0.001
0.01
0.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Dist., m
Conc
., m
g/L
LeftRight
(b)
Figure A.14. Effect of TSCF value on plume concentration for different ET locations.
233
Qin=150, ET at the left edge
-0.05
4.95
9.95
14.95
19.95
24.95
29.95
34.95
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Qin=150, ET at the left edge
1.00
10.00
100.00
1000.00
10000.00
100000.00
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Qin=150, ET at the right edge
-0.05
4.95
9.95
14.95
19.95
24.95
29.95
34.95
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Qin=150, ET at the right edge
1.00
10.00
100.00
1000.00
10000.00
100000.00
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Figure A.15. Average solute mass-flux for different LET lengths and locations, Qin=150 m3/d.
234
Reduction in mass fluxQin=150, ET at the left edge
-0.50
4.50
9.50
14.50
19.50
24.50
29.50
34.50
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d L(ET)/Lp=0.25
L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0NA
Reduction in mass fluxQin=150, ET at the right edge
-0.50
4.50
9.50
14.50
19.50
24.50
29.50
34.50
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d L(ET)/Lp=0.25
L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0NA
Reduction in mass flux
Qin=150, ET at the left edge
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Red
uctio
n in
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Reduction in mass fluxQin=150, ET at the right edge
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Red
uctio
n in
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
% Reduction in mass flux
Qin=150, ET at the left edge
-5
15
35
55
75
95
0 200 400 600 800 1000 1200
Dist., m
% r
educ
tion
in M
ass-
flux
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
% Reduction in mass fluxQin=150, ET at the right edge
-5
15
35
55
75
95
0 200 400 600 800 1000 1200
Dist., m
% r
educ
tion
in M
ass-
flux,
mg/
d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Figure A.16. Average reduction in solute mass-flux (with respect to the NA conditions) for
different LET lengths and locations, Qin=150 m3/d.
235
L(ET)/Lp=0.25
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
LEFTRIGHT
L(ET)/Lp=0.25
1.00
10.00
100.00
1000.00
10000.00
100000.00
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
LEFTRIGHT
L(ET)/Lp=0.50
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
LEFTRIGHT
L(ET)/Lp=0.50
1.00
10.00
100.00
1000.00
10000.00
100000.00
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
LEFTRIGHT
L(ET)/Lp=0.75
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
LEFTRIGHT
L(ET)/Lp=0.75
1.00
10.00
100.00
1000.00
10000.00
100000.00
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
LEFTRIGHT
Figure A.17. Comparison between mass-flux results for different phytoremediation system
dimensions and locations
236
Qin=105
ET at left edge
-0.1
5.0
10.0
15.0
20.0
25.0
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Qin=105ET at left edge
1.0
10.0
100.0
1000.0
10000.0
100000.0
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75
Qin=105
ET at right edge
-0.1
5.0
10.0
15.0
20.0
25.0
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Qin=105 ET at right edge
1.0
10.0
100.0
1000.0
10000.0
100000.0
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75
Figure A.18. Average solute mass-flux for different LET lengths and locations, Qin=105 m3/d.
237
Qin=105, ET at left edge
-1.5
3.5
8.5
13.5
18.5
23.5
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d L(ET)/Lp=0.25
L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0NA
Qin=105, ET at right edge
-1.5
3.5
8.5
13.5
18.5
23.5
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d L(ET)/Lp=0.25
L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0NA
Reduction in mass fluxQin=105, ET at left edge
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Reduction in mass fluxQin=105, ET at right edge
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
0 200 400 600 800 1000 1200
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
% Reduction in mass fluxQin=105, ET at left edge
-100
1020
3040
5060
7080
90100
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
% Reduction in mass fluxQin=105, ET at right edge
-30-20-10
0102030405060708090
100
0 200 400 600 800 1000 1200
Dist., m
Mas
s-flu
x, m
g/d
L(ET)/Lp=0.25L(ET)/Lp=0.50L(ET)/Lp=0.75L(ET)/Lp=1.0
Figure A.19. Average reduction in solute mass-flux (with respect to the NA conditions) for
different LET lengths and locations, Qin=105 m3/d.
238
L(ET)/Lp=0.25
Left edge
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
50.00
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.25Right edge
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.50
Left edge
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
50.00
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.50Right edge
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.75
Left edge
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
50.00
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.75Right edge
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
Figure A.20. Effect of inflow rate on solute mass-flux for different values of LET and ET
locations.
239
L(ET)/Lp=0.25Left edge
-2
-1
0
1
2
3
4
5
6
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Redu
ctio
n in
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.25Right edge
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Redu
ctio
n in
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.50
Left edge
-2
-1
0
1
2
3
4
5
6
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Redu
ctio
n in
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.50Right edge
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Redu
ctio
n in
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.75
Left edge
-2
-1
0
1
2
3
4
5
6
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Redu
ctio
n in
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
L(ET)/Lp=0.75Right edge
-1.5
-1
-0.5
0
0.5
1
1.5
2
0 200 400 600 800 1000 1200 1400
Thou
sand
s
Dist., m
Redu
ctio
n in
Mas
s-flu
x, m
g/d
Qin=200Qin=150Qin=105
Figure A.21. Effect of in-flow rate on the reduction of solute mass-flux (compared to the NA
conditions) for different values of LET and ET locations.
240
L(ET)/Lp=0.25Left edge
-20
0
20
40
60
80
100
0 200 400 600 800 1000 1200 1400
Dist., m
% R
educ
tion
in M
ass-
flux
Qin=200Qin=150Qin=105
L(ET)/Lp=0.25Right edge
-40
-20
0
20
40
60
80
100
0 200 400 600 800 1000 1200 1400
Dist., m
% R
educ
tion
in M
ass-
flux
Qin=200Qin=150Qin=105
L(ET)/Lp=0.50
Left edge
-20
0
20
40
60
80
100
0 200 400 600 800 1000 1200 1400
Dist., m
% R
educ
tion
in M
ass-
flux
Qin=200Qin=150Qin=105
L(ET)/Lp=0.50Right edge
-40
-20
0
20
40
60
80
100
0 200 400 600 800 1000 1200 1400
Dist., m
% R
educ
tion
in M
ass-
flux
Qin=200Qin=150Qin=105
L(ET)/Lp=0.75
Left edge
-20
0
20
40
60
80
100
0 200 400 600 800 1000 1200 1400
Dist., m
% R
educ
tion
in M
ass-
flux
Qin=200Qin=150Qin=105
L(ET)/Lp=0.75Right edge
-20
0
20
40
60
80
100
0 200 400 600 800 1000 1200 1400
Dist., m
% R
educ
tion
in M
ass-
flux
Qin=200Qin=150Qin=105
Figure A.22. Effect of in-flow rate on the percentage reduction of solute mass-flux
(compared to the NA conditions) for different values of LET and ET locations.
241
L(ET)/Lp=0.50
Q(in)=150
-10
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800 1000 1200
Dist., m
% R
educ
tion
in M
ass-
flux
LEFT EdgeRIGHT Edge
L(ET)/Lp=0.25Q(in)=150
-10
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800 1000 1200
Dist., m
% R
educ
tion
in M
ass-
flux
LEFT EdgeRIGHT Edge
L(ET)/Lp=0.75
Q(in)=150
-10
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800 1000 1200
Dist., m
% R
educ
tion
in M
ass-
flux
LEFT EdgeRIGHT Edge
Figure A.23. Effect of ET locations on the percentage reduction of solute mass-flux
(compared to the NA conditions) for different values of LET.
242
L(ET)=0.5Lp (LEFT), t=+3650
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAET
L(ET)=0.5Lp (LEFT),t=+3650
1E-12
1E-11
1E-10
1E-09
1E-08
1E-07
1E-06
1E-05
0.0001
0.001
0.01
0.1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAET
L(ET)=0.5Lp (LEFT), t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
L(ET)=0.5Lp (LEFT),t=+1825
1E-121E-111E-101E-091E-081E-071E-06
1E-050.00010.0010.010.1
1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAET
L(ET)=0.5Lp (RIGHT), t=+3650
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
L(ET)=0.5Lp (RIGHT), t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
Figure A.24. Solute concentration profiles, source removed for LET=0.5Lp at left and right
sides of the plume footprint.
243
L(ET)=Lp, t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NAET
L(ET)=Lp, t=+1825
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NAET
t=+1825
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Con
c., m
g/L
NA L(ET)=0.5Lp (Right) L(ET)=0.5Lp (Left) L(ET)=Lp
t=+1825
1.00E-07
1.00E-06
1.00E-05
1.00E-04
1.00E-03
1.00E-02
1.00E-01
1.00E+00
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NA L(ET)=0.5Lp (Right) L(ET)=0.5Lp (Left) L(ET)=Lp
t=+3650
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NA L(ET)=0.5Lp(Right) L(ET)=0.5Lp(Left) L(ET)=Lp
t=+3650
1E-121E-111E-101E-091E-081E-071E-061E-05
0.00010.0010.010.1
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Conc
., m
g/L
NA L(ET)=0.5Lp(Right) L(ET)=0.5Lp(Left) L(ET)=Lp
Figure A.25. Solute concentration profiles, source removed for LET=Lp, and comparison of
the LET location effect on concentration.
244
L(ET)=0.5Lp at the left edge
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Redu
ctio
n in
con
cent
ratio
n, m
g/L
t=+1825t=+3650
L(ET)=0.5Lp at the right edge
-0.005
-0.004
-0.003
-0.002
-0.001
0
0.001
0.002
0.003
0.004
0.005
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Redu
ctio
n in
con
cent
ratio
n, m
g/L
t=+1825t=+3650
L(ET)=Lp
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Redu
ctio
n in
con
cent
ratio
n, m
g/L
t=+1825t=+3650
t=+1825
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Red
uctio
n in
con
cent
ratio
n,
mg/
L
L(ET)=0.5Lp(LEFT) L(ET)=0.5Lp(RIGHT) L(ET)=Lp
t=+3650
-0.006
-0.005
-0.004
-0.003
-0.002
-0.001
0
0.001
0.002
0.003
0.004
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Red
uctio
n in
con
cent
ratio
n, m
g/L
L(ET)=0.5Lp(LEFT) L(ET)=0.5Lp(RIGHT) L(ET)=Lp
t=+3650
-400
-300
-200
-100
0
100
200
300
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
% r
educ
tion
in C
L(ET)=0.5Lp(LEFT) L(ET)=0.5Lp(Right) L(ET)=Lp
Figure A.26. Reduction in solute concentration (after the source is removed) for different LET
lengths and locations.
245
0
5000
10000
15000
20000
25000
30000
35000
40000
0 730 1460 2190 2920 3650 4380 5110 5840 6570 7300
Time, d
Sol
ute
mas
s, g
0
5000
10000
15000
20000
25000
30000
35000
40000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Sol
ute
mas
s, g
NA (Source removed) ET, source ONNA,Source ON ET, Source removed
(a) (b)
Figure A.27. Solute mass in aquifer after removing the source, (a), and with a
phytoremediation system (b).
0
5
10
15
20
25
30
35
40
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Thou
sand
s
time, d
Sol
ute
mas
s, g
NA, Source ON L(ET)=0.5Lp, Left, Source ONNA(Source removed) L(ET)=0.5Lp, RightL(ET)=0.5Lp, Left L(ET)=Lp
100
1000
10000
100000
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
time, d
Sol
ute
mas
s, g
NA, Source ON L(ET)=0.5Lp, Left, Source ONNA(Source removed) L(ET)=0.5Lp, RightL(ET)=0.5Lp, Left L(ET)=Lp
Reduction in solute mass after using ET
-100
400
900
1400
1900
2400
2900
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
Red
uctio
n in
sol
ute
mas
s, g
LpLeftRight
% reduction in solute mass at different times
-10.0
0.0
10.0
20.0
30.0
40.0
50.0
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650
Time, d
% r
educ
tion
in s
olut
e m
ass
LpLeftRight
Figure A.28. Solute mass reduction due to applying a phytoremediation system where the
contaminant source is removed.
246
-50
0
50
100
150
200
250
300
350
400
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
Mas
s-flu
x, m
g/d
NARightLeftFull
-100
-50
0
50
100
150
0 100 200 300 400 500 600 700 800 900 1000 1100
Dist., m
% r
educ
tion
in m
ass-
flux
FullLeftRight
at X=500
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 100 200 300 400 500
Dist., m
Mas
s flu
x, m
g/d
NAL(ET)=0.5Lp, LeftL(ET)=0.5Lp, RightL(ET)=Lp
at X=1000
-2
0
2
4
6
8
10
12
0 100 200 300 400 500
Dist., m
Mas
s flu
x, m
g/d
NAL(ET)=0.5Lp, LeftL(ET)=0.5Lp, RightL(ET)=Lp
at X=1000
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0 100 200 300 400 500
Dist., m
Mas
s flu
x, m
g/d
L(ET)=Lp
Figure A.29. Solute mass-flux for different ET locations (up), and at downstream cross
sections where the contaminant source is removed.
247
Table A.1. Average mass-flux at different cross-sections downstream the plume source for different values of ET width. W/Ws =3.0, LET=0.5Lp Dist. TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.0
0 31898.91 32641.06 32657.11 32673.36 32689.75 32707.94 100 20398.85 15063.07 15833.02 16644.88 17501.75 18409.24 200 13003.02 6098.404 6722.883 7413.178 8187.245 9042.449 300 8307.07 2002.789 2328.417 2711.859 3162.813 3690.563 400 5318.9 401.407 496.7229 615.74 764.7523 951.1825 500 3413.276 -55.24647 -71.05185 -91.56965 -118.1618 -152.7812 600 2195.318 -7.376162 -9.259253 -11.63977 -14.64152 -18.44344 700 1415.162 -1.085158 -1.276102 -1.504324 -1.776896 -2.103634 800 914.3307 -0.179134 -0.195157 -0.213265 -0.233604 -0.256642 900 592.1101 -0.034383 -0.035371 -0.036421 -0.037584 -0.038831
1000 384.3042 -0.00486 -0.004881 -0.004906 -0.004926 -0.004949 1100 250.0108 -0.001958 -0.001958 -0.001959 -0.001959 -0.001959
W/Ws =2.50, LET=0.5Lp Dist. TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.00
0 31898.91 32915.228 32929.569 32947.587 32965.5517 32983.7527 100 20398.85 15552.211 16329.386 17149.227 18020.208 18934.7625 200 13003.02 6722.3057 7387.622 8128.6091 8946.50284 9852.64807 300 8307.07 2546.0797 2944.0209 3403.6426 3942.1329 4570.15185 400 5318.9 756.91563 924.86608 1132.4308 1389.12725 1705.49563 500 3413.276 108.48747 138.29661 176.59762 225.938574 289.61161 600 2195.318 76.837504 96.320875 120.76384 151.6494 190.654261 700 1415.162 45.506058 54.51716 65.412521 78.6103536 94.6164713 800 914.3307 28.505975 32.031378 36.065089 40.713715 46.0727655 900 592.1101 19.045963 20.135391 21.317288 22.6253461 24.0672915
1000 384.3042 13.072447 13.293012 13.510912 13.7513441 14.0089144 1100 250.0108 8.8791296 8.89893 8.9212269 8.94392278 8.96761858
W/Ws =2.00, LET=0.5Lp Dist. TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.00
0 31898.91 33125.591 33143.185 33159.401 33177.395 33192.463 100 20398.85 15998.211 16784.449 17609.799 18486.468 19409.05 200 13003.02 7327.2167 8032.4327 8804.692 9665.6622 10611.701 300 8307.07 3092.0883 3550.386 4083.0028 4698.708 5412.7932 400 5318.9 1144.5545 1384.4002 1672.1928 2028.033 2459.6353 500 3413.276 324.59511 408.43334 514.84406 650.17981 823.35171 600 2195.318 200.34204 249.02984 309.96671 386.30917 482.0612 700 1415.162 117.72951 142.29053 172.20585 208.66955 253.1439 800 914.3307 73.888286 84.932653 97.830761 112.91809 130.58817 900 592.1101 49.427501 53.720094 58.520344 63.898869 69.925767
1000 384.3042 34.140504 35.41242 36.787762 38.275401 39.885577 1100 250.0108 23.553534 23.790564 24.041694 24.307977 24.587924
248
W/Ws =1.50, LET=0.5Lp TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.00
0 31898.91 33221.3 33238.27 33253.85 33268.05 33285.5 100 20398.85 16415.44 17194.81 18022.01 18893.05 19809.71 200 13003.02 7937.72 8666.178 9462.012 10346.87 11317.6 300 8307.07 3668.934 4176.41 4762.617 5434.548 6211.474 400 5318.9 1580.528 1882.074 2245.309 2678.907 3206.755 500 3413.276 602.8832 743.3309 918.5848 1137.677 1412.275 600 2195.318 371.4798 454.3311 556.6887 683.4037 840.4046 700 1415.162 223.3898 268.5966 323.5199 390.325 471.6692 800 914.3307 142.2885 164.9089 191.563 222.9683 260.1012 900 592.1101 96.01605 106.1652 117.6792 130.7561 145.6247
1000 384.3042 66.67826 70.51732 74.72219 79.33544 84.39885 1100 250.0108 46.44781 47.52271 48.67336 49.90368 51.21418
W/Ws =1.0, LET=0.5Lp TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.00
0 31898.91 31825.68 31829.5 31831.83 31835.62 31839.43 100 20398.85 18154.64 18901.81 19686.67 20516.8 21378.65 200 13003.02 10200.97 10982.25 11838.47 12774.26 13789.77 300 8307.07 5641.729 6281.398 7006.372 7824.646 8754.69 400 5318.9 3083.563 3547.401 4089.315 4734.922 5493.746 500 3413.276 1683.237 1986.419 2354.243 2801.379 3347.478 600 2195.318 985.2151 1155.937 1361.781 1610.492 1911.704 700 1415.162 599.0137 696.4961 812.9197 952.2732 1119.322 800 914.3307 383.5043 436.0639 497.6466 569.8693 654.7882 900 592.1101 258.0699 283.9326 313.4068 347.0149 385.4239
1000 384.3042 178.0829 189.3581 201.8213 215.6152 230.8906 1100 250.0108 123.4365 127.4878 131.8604 136.5642 141.642
W/Ws =3.0, LET=Lp Dist. TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.0
0 31898.91 32321.06 32329.06 32338.59 32348.22 32356.37 100 20398.85 17778.68 18211.04 18657.48 19111.41 19582.24 200 13003.02 9408.858 9845.368 10306.35 10790.78 11301.17 300 8307.07 4774.561 5111.889 5474.634 5863.736 6277.587 400 5318.9 2303.867 2526.054 2769.732 3038.008 3334.301 500 3413.276 1043.815 1173.373 1319.537 1484.582 1670.985 600 2195.318 437.1834 503.9652 581.1245 670.3788 773.6675 700 1415.162 165.9602 195.1042 229.4085 269.7824 317.3085 800 914.3307 54.71013 91.02759 76.92278 91.21343 108.1605 900 592.1101 12.37973 14.69951 17.45394 20.72547 24.61094
1000 384.3042 -2.008 -2.349601 -2.749055 -3.216359 -3.762509 1100 250.0108 -0.328229 -0.371041 -0.419476 -0.474403 -0.536588
249
W/Ws =2.50, LET=Lp Dist. NA TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.00
0 31898.91 32447.57 32457.087 32463.626 32471.7049 32481.393 100 20398.85 18008.547 18433.34 18879.382 19338.5127 19807.9443 200 13003.02 9721.0965 10167.362 10635.449 11127.7879 11644.4855 300 8307.07 5092.5109 5444.9971 5822.509 6224.89563 6662.03361 400 5318.9 2576.2527 2815.7464 3080.7046 3370.85638 3691.18901 500 3413.276 1248.9631 1398.5387 1566.6878 1755.7273 1968.49162 600 2195.318 577.42303 662.09816 759.50282 871.601661 1000.70416 700 1415.162 253.82004 296.81673 347.19879 406.248138 475.455418 800 914.3307 106.14167 125.3787 148.10678 175.00606 206.800709 900 592.1101 40.947532 48.454983 57.349441 67.8908697 80.3803167
1000 384.3042 12.871472 15.056925 17.621116 20.6318845 24.1648052 1100 250.0108 8.7276899 9.8710016 11.170279 12.6478341 14.327968
W/Ws =2.0, LET=Lp Dist. NA TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.00
0 31898.91 32544.76 32552.735 32560.737 32568.806 32578.436 100 20398.85 18210.34 18641.766 19089.232 19544.285 20016.844 200 13003.02 10021.889 10471.943 10952.586 11442.253 11965.426 300 8307.07 5409.2292 5772.2386 6161.5756 6577.9675 7024.084 400 5318.9 2850.1809 3108.3195 3389.3483 3697.7831 4034.8452 500 3413.276 1463.2067 1631.0019 1818.7869 2029.1053 2265.9983 600 2195.318 729.95891 831.68738 948.06837 1081.3224 1233.8486 700 1415.162 354.96645 412.09014 478.63081 556.14049 646.47863 800 914.3307 169.89475 199.27628 233.85835 274.54873 322.44203 900 592.1101 79.821713 93.774732 110.23808 129.67161 152.60877
1000 384.3042 35.660181 41.480425 48.302242 56.299961 65.676443 1100 250.0108 23.262271 26.256892 29.667631 33.551904 37.974631
W/Ws =1.50, LET=Lp Dist., m NA TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.0
0 31898.91 32586.05 32592.2 32599.87 32609.15 32616.94 100 20398.85 18407.35 18831.02 19272.13 19723.59 20188.23 200 13003.02 10349.2 10788.42 11261.3 11751.7 12271.11 300 8307.07 5753.273 6119.064 6514.218 6922.434 7370.956 400 5318.9 3156.064 3422.757 3715.079 4027.64 4374.168 500 3413.276 1706.277 1886.314 2086.579 2309.464 2555.937 600 2195.318 908.8326 1023.89 1154.434 1302.64 1470.946 700 1415.162 478.9115 548.3149 628.3736 720.8272 827.6331 800 914.3307 252.5642 291.7524 337.4098 390.6813 452.8993 900 592.1101 133.8801 154.6711 179.007 207.5052 240.8943
1000 384.3042 70.25268 80.37786 92.16507 105.8996 121.9102 1100 250.0108 46.03622 51.36535 57.41988 64.30616 72.13453
250
W/Ws =1.0, LET=Lp Dist., m NA TSCF=1.0 TSCF=0.75 TSCF=0.50 TSCF=0.25 TSCF=0.0
0 31898.91 32501.35 32503.23 32505.13 32505.51 32507.38 100 20398.85 18735.71 19111.17 19492.47 19889.82 20287.22 200 13003.02 10850.77 11243.68 11657.94 12085.06 12534.95 300 8307.07 6261.872 6591.792 6935.06 7310.616 7699.612 400 5318.9 3598.162 3840.802 4102.287 4390.332 4700.445 500 3413.276 2053.235 2223.975 2411.615 2618.137 2845.316 600 2195.318 1165.472 1279.646 1407.207 1549.955 1709.73 700 1415.162 660.3766 733.3096 816.0255 909.9049 1016.544 800 914.3307 376.9442 421.0981 471.6537 529.5853 596.082 900 592.1101 217.8663 243.1854 272.3007 305.8381 344.4571
1000 384.3042 126.4624 140.0181 155.55 173.3812 193.8649 1100 250.0108 83.15738 90.45315 98.6584 107.9001 118.313
251
VITA
Amr A. El-Sayed was born on August 27, 1968, in El-Minia, Egypt. He finished High School in 1986, and he was ranked first in the mathematical branch for his district, as the high school examination is the same nationwide in Egypt. In 1986, he entered the program of Faculty of Engineering, El-Minia University, Egypt. During his study in the Engineering school, he was elected the ideal student for El-Minia University, Egypt in the year 1988, and was elected the Ideal student for dorms three years in a row. El-Sayed, Amr graduated in 1990 from the civil engineering department, and he was ranked first with the degree of honor. He started his graduate studies in the year 1992 as a TA/RA in the civil engineering department, faculty of Engineering, El-Minia University, and had his Master degree in the year 1996 under the title “Effect of non-homogenous layers beneath floors of hydraulic structures on seepage characteristics”. During the period of 1994 to 2000, El-Sayed was working as part-time designer in many consulting offices. In the year 1999, he was awarded a scholarship from the Egyptian government to pursue his Ph. D. degree in civil engineering. He entered Virginia Tech civil and environmental engineering program in the Fall of 2000 and graduated in the Fall of 2006.
252
AMR A. EL-SAYED
310D Patton Hall, Virginia Tech., Blacksburg VA 24060 (540) 257-4192/(540) 231-4421 E-mail: [email protected], webpage: http://www.filebox.vt.edu/users/aelsayed/amr.htm
EDUCATION Virginia Polytechnic Institute and State University, VA
PhD Candidate (Environmental & Water Resources), graduating Dec 06 El-Minia University, Egypt, El-Minia, Egypt, 1996
Master of Science in Civil Engineering Thesis title: "Effect of non-homogenous layers beneath floors of hydraulic structures on seepage characteristics"
El-Minia University, Egypt, El-Minia, Egypt, 1990 B. Sc. in Civil Engineering Rank 1st. With the degree of Honor
KNOWLEDGE/SKILLS Numerical Modeling for groundwater/contaminant transport: GMS 6.0,
FORTRAN Land Developments/Road Design using Autodesk Products: Land Desktop
2007, Civil 3D 2007, Revit Building 8.1, MicroStation, and GEOPAK Surface water simulation/Pipe Network analysis using: HEC-HMS & HEC-
RAS, Bentley (Haestad Methods): WaterCAD, StormCAD Geospatial Analysis using ArcGIS 9.x
ACCOMPLISHMENTS 1- More than 14 years experience in civil engineering. 2- Re-planning main roads leading to Cairo Stadium 1990. 3- Design of many buildings in El-Minia University, Egypt (List of projects is
available at request). 4- Design the webpage, and CD of EWRI conference in Roanoke, VA in
2002 (Reference: Prof. David Kibler, Virginia Tech) 5- Create digital CAD files for the "Slope Stability Manual" U.S Army's corps
of Engineers. (Reference: Prof. Mike Duncan, Virginia Tech) 6- SEAM3D – Plant Uptake Package (PUP), Technical report to USGS.
(Reference: Prof. Mark Widdowson, Virginia Tech)
EMPLOYMENT HISTORY Virginia Polytechnic Institute and State University, Blacksburg VA Ph. D. student/TA/RA/Instructor
• Fall 2005, Spring 05, and Fall 2006: Instructor and course builder for CEE 4204 CAD Applications in CEE.
• Spring 2005: Instructor EngE 1234 Engineering Hands-on Lab.
• Fall 2004: Instructor EngE 1024 Engineering Exploration.
• Spring 2004: Instructor CEE 3304 Fluid Mechanics.
• Summer 2003: TA CEE 3304 Fluid Mechanics. • Spring 2002, Fall 2002, Spring 2003: Instructor
EF 1234 Hands-on Lab. • Fall 2001: TA CEE 4334 Hydraulic Structures. • Summer 2001: TA CEE 3304 Fluid Mechanics
(08/00 - present)
253
EMPLOYMENT HISTORY El-Minia University, El-Minia, Egypt Research Assistant/M. Sc. Student
(01/92 - 08/00)
Engineering consultation office (ECO), El-Minia, Egypt Designer engineer Software developer, prepare CAD drawings, and technical reports.
(01/92 - 08/00)
Road construction company, Cairo, Egypt, Site engineer Road construction engineer.
(01-90 - 01/92)
RELATED COURSEWORK
Fluid Mechanics for CEE Water Resources and Hydrology. CAD/GIS Applications in Civil and Environmental Engineering. FORTRAN/Visual BASIC programming.
PRESENTATIONS/
PROJECTS
Water surface simulation using HEC-RAS (Guest speaker for CEE 3314_Water Resources Engineering, Spring 2006)
Increase your productivity: Autodesk Civil 3D. (Seminar guest speaker in Auburn Univ., AL, Fall 2006)
Watershed Delineation, and Travel Time Calculation using GIS (Guest speaker for CEE 5224_Advanced GIS, Fall 2004)
Strouble’s Subwatershed Study Area (Storm Sewers Project for CEE 5224, Fall 2002)
Design of Earth Dams using SEEP2D (Guest speaker for CCEE 4334_Hydraulic Structures, Fall 2001)
INTERESTS Fine Arts, Reading Table Tennis, Basketball Traveling/interaction with people.
KEYWORD SUMMARY Civil Engineering/Education CAD/GIS applications Water recourses/Groundwater Modeling