Home

Probabilistic Remediation Evaluation Model for - CLU-IN

1. 23 PEME 7 dpa GE 23 23 View Graphical or tabular OWDDUE 23 _________ 24 Model Input Variables 44 DNAPL Source Parameters and Dimensions 44 Source Parameters Initial Source 0 ecccceeecccceseccceesececeeccseeseceeeeeeceseeees 44 SOUECS Soi aa E E 45 Transport Parameters Spo eiie tn A 47 Plume Decay Rates and Yield Coefficients ra eta en e usato e fus 260 49 49 Plume REACH on 85952 0 faites eia N 51 Plume Species First Order Decay Rates ccccccccccccceccceeeceeeeeeseeeeeeceeeeeeeeseeeaaaaaas 51 Plume Treatment Dimensions and 5 5 22 Cancer Risk 53 Tutorials 55 E 55 Tutorial 1 Deterministic Thermal Remediation of a PCE 56 Manual for PREMChlor Draft Contents e iii Tutorial 2 Thermal Remediation of a PCE Source Considering Uncertainties in Source terse achat ore Ae ee 59 Tutorial 3 Probabilistic Thermal Remediation of a PCE Source
2. MQQQQAY RQ PREMChlor provides many intermediate and final outputs Eighteen useful final outputs can be accessed after simulation PREMChlor considers up to four parent daughter compounds These results are assembled into the probability statistics and the probability distribution Probability statistics include time among all of the realizations From the median line upward the outline of the light dot filled area 1 the 75 percentile the outline of light diagonal filled area is the 95 percentile and the outline of dark dot filled area 1s the upper bound From the median line downward the outline of the light dot filled area is percentile of 128 ug L 25 of 419 ug L median of 647 ug L 75 of 929 ug L 95 of 1185 ug L and the during natural attenuation is shown in Figure 6 The x axis is the simulation duration time and the y axis is the concentration The concentration shown here is the plume centerline mass concentration and the the 25 percentile the outline of light diagonal filled area is the 5 percentile and the outline of dark dot filled area 1s the lower bound The probabilistic statistics also are displayed in tabular form Figure 7 In this natural attenuation example the concentration at the 30 yr has the lower bound of 49 ug L 5 upper bound of 1337 ug L The probability statistics of an output are displayed by the time histories in the form of the probability histories An example of
3. 7 0 01 om 00001 go back to the main interface click the Main Interface tab jump to another interface click the Go tab on the Run Controller and select an interface Manual for PREMChlor Contents e 28 6 Click on the Source Remediation tab and the source zone remediation parameters screen will appear source plume function SourceRemediation i 5 Ioj x Source Remediation If checked use StartTime End Time Min Likely Deterministic Value Remediatinon Time tyr 30 31 Aqueous Phase Source Decay 1yr 0 01 0 1 m 0 01 Percent Removed Unit Cost cubie meter If checked use If checked use Mean Stdw Max Deterministic Value Mean ody Max Deterministic value V Thermal Methods Tolo E D 0957 PT SED scies z 151 SurfactantFlooding 05 00 onf 1 r os 50356 375 3632 71937 50356 BE oo 16349 65 2616 66752 16349 7 Bioremediation oss oof 1 Pl 055 3793 sf 262 29429 3793 lf checked use enhanced aqueous phase decay rate 0 8 0 6 0 95 0 8 for remediation efficiency source Containment Kain Interface To go back to the main interface click the Main Interface tab To jump to another interface click the Go tab on the Run Controller and select an interface Manual for PREMChlor Contents e 29 fe Click on the Plume Decay Rates tab and the following screen will appear This screen
4. eene 62 Tutorial 4 Probabilistic Thermal Remediation of PCE Source with Enhanced Plume tt tree rt 65 References 72 Manual for PREMChlor Draft Contents e iv Acronym List EPA CVOC DCE DLL DNAPL GUI LB NAPLs NPV NRC O amp M PAT PCE PDFs PRBs REMChlor TCE UB VC Environmental Protection Agency Chlorinated volatile organic compound dichloroethylene Dynamic Link Library dense non aqueouse phase liquid graphical user interface Lower Bound Non aqueous phase liquid prestent net value National Research Council Operation amp Management pump and treat tetrachloroethylene Probability density functions Permeable reactive barriers Remediation Evaluation Model for Chlorinated solvents trichloroethylene Upper bound vinyl chloride Manual for PREMChlor Draft Contents e v Overview Abstract A new probabilistic remediation model Probabilistic Remediation Evaluation Model for Chlorinated solvents PREMChlor has been developed for simultaneously evaluating the effectiveness of source and plume remediation considering the uncertainties in all major parameters This development was conducted as an Environmental Security Technology Certification Program ESTCP research project ER 0704 which was a joint effort between Clemson University GSI Environmental Inc and Purdue University The technical foundation of PREMChlor 1 the U S Env
5. Percentiles 451731 453454 452807 458532 4800867 433055 506369 517r6 523608 532346 534440 Statistics Mean 4 9S02E 5 Mean Cont 5795 4 S83BE 5 4 5 5 D 18413 Skewness 0 010499 2 479 Mum Reals 100 26 05 2 05 1 5 05 1 0 05 5 0 06 1 E 5 05 PDF Confidence Bounds Show Marker Calculator Cum Probability 9 5 E Value 7 550 3E 5 513181 Probability Density Cond Tail Expectation Result Array Close Manual for PREMChlor Contents e 71 References Alvarez P J J and W A Illman 2006 Bioremediation and Natural Attenuation Hoboken New Jersey Wiley Interscience Aziz C E CJ Newell J R Gonzales P Hass Clement and Sun 2000 BIOCHLOR Natural Attenuation Decision Support System User s Manual Version 1 0 U S Environmental Protection Agency USEPA 600 R 00 008 Brusseau M L D A Sabatini J S Gierke and M D Annable ed 1999 Innovative Subsurface Remediation Field Testing of Physical Chemical and Characterization Technologies ACS Symposium Series 725 Washington D C American Chemical Society Chang S S 1999 Implementing Probabilistic Risk Assessment in USUSEPA Superfund Program Human and Ecological Risk Assessment An International Journal Vol 5 737 754 Delshad M G A Pope and K Sepehrnoori 1996 A Compositional Simulator for Modeling Surfactant Enhan
6. USER S GUIDE PREMChlor Probabilistic Remediation Evaluation Model for Chlorinated Solvents ESTCP Project ER 0704 MARCH 2010 Hailian Liang Ph D Ronald Falta Ph D Clemson University Charles Newell Ph D Shahla Farhat Ph D GSI Environmental Inc P Suresh C Rao Ph D Nandita Basu Ph D Purdue University Distribution Statement A Approved for Public Release Distribution is Unlimited Environmental Security Technology Certification Program Form Approved Report Documentation Page Public reporting burden for the collection of information is estimated to average hour per response including the time for reviewing instructions searching existing data sources gathering and maintaining the data needed and completing and reviewing the collection of information Send comments regarding this burden estimate or any other aspect of this collection of information including suggestions for reducing this burden to Washington Headquarters Services Directorate for Information Operations and Reports 1215 Jefferson Davis Highway Suite 1204 Arlington VA 22202 4302 Respondents should be aware that notwithstanding any other provision of law no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number 1 REPORT DATE 3 DATES COVERED MAR 2010 00 00 2010 00 00 2010 4 TITLE AND SUBTITLE 5a CONTRACT NUMBER PREMChlor
7. Probabilistic Model Predicted Value Time or distance Figure 3 Illustration of deterministic a and probabilistic b modeling approaches Manual for PREMChlor Contents e 15 number of times Each simulation is independent and equally likely referred to as a realization of the system For each realization all of the uncertain parameters are sampled from the specified distributions The performance of the system from one realization is then computed or evaluated and the result 15 saved After repeating many realizations the results of the independent simulations are assembled into probability statistics and distributions A probabilistic simulation consists of hundreds or thousands of deterministic Monte Carlo realizations As illustrated in Figure 4 during the probabilistic simulation GoldSim is used to specify the probability distributions for all stochastic parameters and specify the Monte Carlo parameters such as the total simulation duration time step and the total realization number for the probabilistic simulation Inside the Monte Carlo loop for each realization GoldSim is used to sample the value for each uncertain parameter through its PDF and specify the value to each deterministic parameter and assigns the values to REMChlor The REMChlor FORTRAN source code is called via a FORTRAN DLL application to perform the analytical calculation and calculation results are passed back to GoldSim After all of the realizations are
8. 2016 122 41 8 5463 B7 258 103 05 275 21 369 72 Jul 02 2017 120 43 7 3844 85 742 101 39 271 55 363 37 Jul 02 2018 119 59 7 2855 5 491 10102 269 95 358 24 02 2019 119 06 7 2838 65 366 1007 268 07 353 25 Jul 01 2020 118 43 7 2807 55 242 100 38 255 B8 348 28 Jul 02 2021 117 81 72777 65 117 100 05 265 47 343 35 Jul 02 2022 11748 72747 54 994 99 729 284 25 338 45 Jul 02 2023 116 57 TOUT 54 87 38 407 263 03 333 57 Jul 01 2024 115 95 7 2687 99085 15505 261 82 115 34 7 2657 64 624 38 765 260 62 323 92 02 2026 114 73 7 2628 B4 501 38 446 259 41 319 14 Jul 02 2027 11412 7 2598 B4 378 98 127 25821 314 38 Jul 01 2028 113 52 7 2568 4 256 oF 81 257 02 309 57 gt Manual for PREMChlor Contents e 61 Tutorial 3 Probabilistic Thermal Remediation of a PCE Source Manual for PREMChlor This tutorial example is identical to the previous source remediation case except that we will now make one more parameter source removal fraction uncertain and run a probabilistic simulation of a source remediation The remediation concentration goal remains same Double click the tutorial file Tutorial3 gsp to open it and then click on the Simulation Settings tab from the main interface Under the Time tab the time settings show the simulation duration time is set as from 1985 to 2085 with a time step of 1 year under the Monte Carlo tab the sim
9. B B 4 gt Container Path Model mains Concentration results v Search Options Sy Model Time Histories of Concentrations E Model main Bf CancerRisk results Concentration results 2 oe cone E concent ee concent dist conctot concen concern concens concend concenz ry concen _dist E concen3_dist qMM concent Distributions of Concentrations Lx concen4_dist gu conctok 0 es conctot_dist ij Cost results it if Discharge results 5 Interfaces conctot dist concen dist concenz dist concena dist concen4 dist lt li i Containment P Class View Result Made Scale 100 Filter OFF Manual for PREMChlor Contents e 40 Manual for PREMChlor The top row elements are the time histories of concentrations for the toal conctot component 1 concen1 component 2 concen2 component 3 concen3 component 4 concen4 and all four components conc Double click the element concen1 to show the probabilistic time history of the component 1 pper Bound 95 Percentile 5595 Percentile 25 Percentile 5 Percentile i Boung Median Repeat steps 11 16 to view other displays for the concentration of the component 1 The bottom row elements are the distributions of concentrations for the toal conctot_dist component 1 concenl_dist component 2 concen2_dist compo
10. Park and Parker 2005 suggest values greater than 1 for finger dominated residual DNAPL and less than 1 for DNAPL pools Essentially should be considered as an uncertain parameter whose mean value can be roughly estimated but whose actual value may never be precisely known at a site Manual for PREMChlor Contents e 11 The solution of Equation 1 with the power function Equation 2 can be used to predict the time dependent depletion of the source zone mass by dissolution The time dependent mass is then used in Equation 2 to calculate the time dependent source discharge If is constant the solutions are given by Falta et al 2005b 1 1 1 CCo Mj 3 A M A M 1 Tan NT E CE yn 4 0 5 0 5 0 This source model can account for aggressive source remediation efforts such as excavation thermal treatment alcohol or surfactant flooding or chemical oxidation that remove a certain fraction of the source mass over a short period of time Falta et al 2005a By rescaling the equations following the removal of source mass the source mass and source discharge due to source remediation are presented by Falta et al 2005b as 1 i mi 6 B gt M 6 2 M 1 X M 7 8 0 g 0 where t is the time when the remediation ends M is the source mass before remediation and 15 the source mass at tz X is the fraction o
11. Probabilistic Remediation Evaluation Model for Chlorinated Solvents 5b GRANT NUMBER 5c PROGRAM ELEMENT NUMBER 6 AUTHOR S 5d PROJECT NUMBER 5 TASK NUMBER 5f WORK UNIT NUMBER 7 PERFORMING ORGANIZATION NAME S AND ADDRESS ES 8 PERFORMING ORGANIZATION Clemson University Clemson SC 29634 REPORT NUMBER 9 SPONSORING MONITORING AGENCY NAME S AND ADDRESS ES 10 SPONSOR MONITOR S ACRONYM S 11 SPONSOR MONITOR S REPORT NUMBER S 12 DISTRIBUTION AVAILABILITY STATEMENT Approved for public release distribution unlimited 13 SUPPLEMENTARY NOTES 14 ABSTRACT A new probabilistic remediation model Probabilistic Remediation Evaluation Model for Chlorinated solvents PREMChlor has been developed for simultaneously evaluating the effectiveness of source and plume remediation considering the uncertainties in all major parameters This development was conducted as an Environmental Security Technology Certification Program ESTCP research project ER 0704 which was a joint effort between Clemson University GSI Environmental Inc and Purdue University The technical foundation of PREMChlor is the U S Environmental Protection Agency USUSEPA REMChlor for Remediation Evaluation Model for Chlorinated solvents model Falta 2008 REMChlor is a significant improvement on existing chlorinated solvent transport models because it can simultaneously account for both source and plume remediation REM Chlor includes a source
12. Remediation interface click the Treatment Dimensions Costs tab and the Plume Treatment interface will appear The Plume Treatment interface shows that the treatment zone Zone in Period 2 has a Treatment Width of 30 m and a Treatment Depth of 5 m The Unit Cost in the treatment zone Zone in Period 2 is made uncertain and has a triangular distribution with a Min of 1 a Likely of 2 and a Max of 3 m The another treatment zone Zone 2 in Period 2 is remaining the natural attenuation condition so dimensions and the unit cost for this zone does contribute to the remediation cost The Present Annual O amp M Cost has a deterministic value of 10 000 The Inflation Rate is 4 and the Interest Rate is 6 The Treatment Period is 75 years source plume function Plume Treatment Plume Treatment Natural Attenuation Natural Attenuation Natural Attenuation Period 3 Treatment one Treatment one Treatment width m 30 Treatment Width 50 Treatment Depth 5 Treatment Depth m 5 Unit Cost S cubic meter Unit Cast icubic meter If checked use Ift checked use Min Likely Max Deterministic value Min Likely Deterministic value ap Matural Attenuation Period 2 in m A E i tplume 1 Natural Attenuation Natural Attenuation Natural Attenuation 1 v1 300 Distance From Source Meters
13. Repeat Sampling Sequences Random Seed Deterministic Simulation Solve Simulation deberministically using Element Deterministic Values Bement Mean Values C Specified Quantile 8 5 Manual for PREMChlor Contents e 25 3 Click on the Explore Model tab and the following screen showing top level of the model stucture will appear Ed GS Player sourceplume thist3 dash4c risk4 cost2la triangular File View Model Help 93 amp ies miae name Search Options Model 2 Model main 9 8 CancerRisk results 8 Concentration results 9 08 Cost results 9 8 Discharge results E Interfaces 8 Componenti EE Component1_Rem Eg Component2 le Component2 Rem EE Component3 EE Component3_Rem EE Component4 Component4_Rem Main 8 Plume Treatment Source_Remediation SourceParameters TransportParameters Remediation Transport_Model Gi Observation_Location Plume DecayRates 2 Run Properties 3 ElapsedTime SlopeFactor Yield m Ga Containment Run Mode Press F5 to start simulation Scale 100 iter OFF 7 Interfaces Remediation ee 29985 Cost results 3 i v 4 This model is orgnized in a top down manner The level of detail increases as the user push down into the container hierarchy For example click the plus sign on the top left side of the Transport_Model container the struc
14. 6 Log normal distribution Beta distribution Figure 5 Distributions used for input parameters in PREMChlor Manual for PREMChlor Contents e 17 A percentile is the the mass discharge of each 50 percentile also known as and the cancer risk posed by each component and the total mass discharge and cancer risk are the commonly used metrics 9 All these values and observations respectively 9 and the total remediation cost 9 the changes of concentrations mass 2 Lower bound LB and upper bound UB the m The solid line is the median of the concentration over the 2 9 the plume remediation cost 2 Concentration vs Time 100 0 z 0 and different percentiles 100 m yz0 m 7 MQQOQQQQQQQQQQLLY MRO QQ MOO QQ MOO QQ NNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNN MQ NNNNNNNNNNNNNNNNNN MY NNNNNNNNNNNNNN MQ QQQOQQQQQ MQ QQQQQQQ QQ MN MOO QQ AL MRO OQ MQ YL MRO QQ NNNNNNNNNNNNNNNNN MRO QQ NNNNNNNNNNNNNNN NNNNNNNNNNNNNNN NNNNNNNNNNNNNNN NNNNNNNNNNNNNNN MOO QQ QOL NNNNNNNNNNN MOQ MQQQQQQY MW QQ Time ANNNNNNNNNN 0 4446 NNNNNNNNNN WAN
15. PREMChlor Description Background Contamination of groundwater by dense non aqueous phase liquids DNAPLs such as chlorinated solvents poses a widespread and serious threat to groundwater supplies due to their toxicity While the solubilities of DNAPLs are very low they are typically several orders of magnitude higher than drinking water standards Pankow et al 1996 DNAPLs tend to act as continuous long term sources of eroundwater contamination and form large dissolved contaminant plumes Technologies have been developed for both DNAPL source control and plume treatment Source control includes either removal or destruction of the contaminant source or its physical isolation For chlorinated solvent source remediation in situ technologies include thermal methods e g steam flooding and electrical heating chemical oxidation surfactant flooding and cosolvent flooding soil vapor extraction and air sparging Reddi 1996 Brusseau et al 1999 Kaluarachchi 2001 USEPA 2004b Mayer and Hassanizadeh 2005 To prevent or reduce the source contaminant loading to the plume source containment methods such as slurry walls clay caps and sealable joint sheet pile walls can be used for isolating the contaminant source Chlorinated solvents in dissolved plumes can be removed by pump and treat PAT systems by natural attenuation processes including biodegradation processes or they can be controlled by reactive barriers Currently enhanced in
16. Present Annual O amp M Cost Treatment Period 75 ftolume tolumed Treatment Rate If checked use Min Likely Max Deterministic Value Inflation Rate 4 TERT ain Interface 1000 100000 16 10000 IntersstRate Manual for PREMChlor Contents e 68 After viewing the parameter settings go back to the main interface and click the Run Model tab After simulation is completed select Total Concentration from the main interface to view the probabilistic result of the concentration 100 meters downgradient from the source Mi Time History 1 0e04 1 0 03 1 0 02 1 0 01 Concentration 1 0 00 1995 2005 2015 2025 2035 Upper Bound ji uum 759 Percentile 25 Percentile 5 LmwerBaund Wean EH Time History AGE Date Time Lower Bound 5 Median 95 Upper Bound A Jul 02 2014 147 15 341 40 473 141 48 370 2 537 05 Jul 02 2015 75 761 31187 6 5392 62 611 207 12 292 82 Jul 01 2016 71 389 25541 6 0057 60 302 201 86 28357 Jul 02 2017 70 875 25345 5 8597 58 963 200 29 281 42 Jul 02 2018 70 537 2 5285 5 8528 59 782 198 33 278 58 Jul 02 2019 70 211 2523 5 8431 58 605 198 42 277 78 Jul 01 2020 63 886 25175 58333 58 41 197 51 275 98 Jul 02 202 69 563 2512 5 8236 59191 196 64 27419 Jul 02 2022 58 241 25065 5 8138 58 974 195 7 2724 Jul 02 2023 68 921 25011 5 8041 58 728 194 82 270 62 Jul 01 2024 88 602 2 4856 5 7945 58 477 193 93 25
17. There are 86 input parameters in PREMChlor Among those 18 are treated as deterministic and 68 are treated as stochastic Deterministic parameters usually have less or no variability and can be defined in a certain way For example two parameters the times when remediation starts and ends are treated as deterministic because they are known parameters for a remediation design Stochastic parameters are normally associated with much uncertainty For example groundwater Darcy velocity is treated as stochastic parameter Stochastic input parameters are defined by distributions During the probabilistic simulation PREMChlor samples the different values from the distribution for each realization In PREMChlor distributions include the triangular distribution normal distribution log normal distribution beta distribution Figure 5 A triangular distribution is defined by a minimum value a most likely value and a maximum value A normal distribution is defined by a mean and a standard deviation A log normal distribution is defined by a geometric mean and a geometric standard deviation A beta distribution is defined by a minimum value a mean a standard deviation and a maximum value More information about these distributions can be found in GoldSim User s Guide 2007 0 30 0 20 2 3 4 5 6 7 8 9 10 O 2 4 6 8 10 12 14 16 18 20 Triangular distribution Normal distribution 1 2 3 4 5
18. al 1996 and BIOCHLOR Aziz et al 2000 simulate remediation by natural attenuation of dissolved hydrocarbons at petroleum fuel release sites and dissolved solvents at chlorinated solvent release sites respectively Several three dimensional Manual for PREMChlor Contents e 9 multiphase numerical models focus on the source zone behavior such as T2VOC Falta et al 1992 and UTCHEM Pope and Nelson 1978 Delshad et al 1996 Recently a new analytical screening level model REMChlor has been developed Falta et al 2005ab and Falta 2008 REMChlor is a significant improvement on existing analytical chlorinated solvent transport models because it can simultaneously account for both source and plume remediation REMChlor includes a source model based on a power function relationship linking the source mass to the source discharge and an analytical plume model based on one dimensional advection with three dimensional dispersion Process and parameter uncertainty that occurs in source and plume remediation is a key factor that has made decision making about DNAPL site remediation alternatives difficult Uncertainties arise from hydrogeological and biogeochemical properties e g hydraulic conductivity from the site condition and history e g size and timing of contaminant releases and discharge to groundwater from the effectiveness of remediation e g fraction of source removed and from the cost of remediation The conventional determinis
19. amp Management Tools DNAPL Sites Probabilistic Remediation Evaluation Model for Chlorinated Solvents Considering Uncertainty Observation Location X m Y mi Z m Results Number of Stream Tube 100 REALIZATION I a ELAFSED m __ TIHE STEF 0 LU SIHULATIOH TIHE aR m L E BASE Manual for PREMChlor Contents e 24 2 Click on the Simulation Settings tab and the following screen will appear This Time setting screen shows the Basic Time Settings including simulation duration and the Time Phase Settings including time steps Simulation Settings EF x Time Monte Carlo Globals Information Specify model start time and duration and define global time steps For model calculations and result plotting Basic Time Settings Time Display Units ro G Duration I Start time 7 2j2001 12 00 00 End time 2172008 100 00 Time Phase Settings Click the Monte Carlo tab to view the Monte Carlo options Click OK to go back to the main interface Simulation Settings Time Monte Carlo Globals Information Define Monte Carlo options to carry out a probabilistic simulation and specify the sampling method For Stochastic variables Realizations 100 Histories to save Bun the following Realization only Realization v Use Latin Hypercube Sampling
20. main Time Histories of Mass Discharges 88 CancerRisk results 8 Concentration results ES cone WV E concent wie we La concent _dist be concen dischtat disch disch2 disch3 discha discharges concenz dist concen concen3 dist concent Distribution of Mass Discharges concen4_dist E canckat conctot disk Ef Cost results 7 fN fN fN T 8 Discharge results sii oll lel dischtot dist dischl dist disch2 dist disch3 dist disch4 dist E Class View 5 Scale 100 Fiter OFF The top row elements are the time histories of mass discharges for the toal dischtot component 1 disch1 component 2 disch2 component 3 disch3 component 4 disch4 and all four components discharges The bottom row elements are the distributions of mass discharges for the toal dischtot _ dist component 1 dischl dist component 2 disch2 dist component disch3 dist and component 4 disch4 dist Click an element to view the mass discharge for a single component To go back to the main interface click the Go tab on the Run Controller and select Main Manual for PREMChlor Contents 42 20 To view the cancer risk for each component click the plus sign on the top left side of the CancerRisk_results container from the top level of model structure window shown in step 18 and the following window will appear Ed GS Player REMChlo
21. model based on a power function relationship linking the source mass to the source discharge and an analytical plume model based on one dimensional advection with three dimensional dispersion The plume model simulates natural attenuation or plume remediation for parent and daughter compounds in the first order sequential decay chain Plume model also calculates the cancer risks posed by carcinogenic compounds assuming that the contaminated water is used in a house for drinking bathing and other household uses PREM Chlor is developed by linking the analytical model REMChlor to a Monte Carlo modeling package GoldSim via a FORTRAN Dynamic Link Library DLL application In PREMChlor all of the uncertain input parameters are treated as stochastic parameters represented by probability density functions PDFs The outputs from PREMChlor are probability distributions and summary statistics of the distributions Cost analysis of common technologies for dense non aqueouse phase liquid DNAPL source removal and dissolved plume treatment are included PREMChlor gives users a single platform where cost source treatment plume management monitored natural attenuation and risk assessment can all be evaluated together and where uncertainty can be incorporated into the site decision making process A license free file containing the user friendly graphical user interface GUI has been generated to make PREMChlor available for use by others 15 SUBJECT TERM
22. risk assessments Oregon DUSEPArtment of Environmental Quality Portland OR Interstate Technology amp Regulatory Council ITRC 2006 Life Cycle Cost Analysis RPO 2 Washington D C pp4 Manual for PREMChlor Contents 72 Jawitz J W A D G G Demmy 5 Berglund and P S C Rao 2005 Groundwater contaminant flux reduction resulting from nonaqueous phase liquid mass reduction Water Resources Research Vol 41 no 10 W10408 Kaluarachchi J J ed 2001 Groundwater Contamination by Organic Pollutants Analysis and Remediation ASCE Manuals and Reports on Engineering Practice No 100 Reston Virginia American Society of Civil Engineers Li H G Huang G Zeng I Maqsood Huang 2007 An integrated fuzzy stochastic modeling approach for risk assessment of groundwater contamination Journal of Environmental Management Vol 82 173 188 Liang H 2009 Probabilistic Remediation Evaluation Model for Chlorinated Solvents Considering Uncertainty PhD Dissertation Dept of Environmental Engineering and Earth Science Clemson University Clemson SC Liu L S Y Cheng and H C Guo 2004 A simulation assessment modeling approach for analyzing environmental risks of groundwater contamination at waste landfill sites Human and Ecological Risk Assessment An International Journal Vol 10 2 373 388 Mayer A S and S M Hassanizadeh ed 2005 Soil and Groundwater Contamination Nonaqueous Phase Liquids
23. the probability histories display of concentration vs time at a location x y z PREMChlor allows users to specify such a location by entering any x y and z value The final results also the median is the value below which 50 percent of the observations may be found Such probability discharges and cancer risks over time time histories are calculated for a specified location x y z statics are useful to evaluate the remediation alternatives include the concentration of each component and the total concentration component and the total mass discharge cancer risk Contaminant concentration value of an output below which a certain percent of observations fall The to assess the performance of the remediation In PREMChlor Each output has multiple values computed from different realizations lowest and highest values for an output among all of the realizations include the source remediation cost the lower and upper bounds location is defined by x Contents 18 Graphic view Figure 6 Probability histories of an output Manual for PREMChlor 88 Time History T A cU ES be Time 5 2 n 0 0 o 2428 79 12477 10802 ___48588 _ 6634 _10078 36434 539 82 52963 15416 _ 780 7 115 85 Median 426 1247 7 1192 1 1076 368 B7 87214 429 99 432 96 1295 1 1318 2 1271 1 1342 8 1332 1190 3 1297 2 1372 6 11121 1250 4 1350 5 1039 1205 785 25 970 84 1161 3 1
24. toEal x his e LA Le n ka T Percentiles 2 5e 05 CumProb ____ ____ 2 De n5 0 001 470456 0 01 474820 0 05 487936 0 1 498559 1 06 05 0 25 541433 o5 S528681 5 06 0 75 542658 _ 0 9 555639 288 0 95 561806 5 5 5 5 5 0 99 587299 POF 0 999 592134 1 56 05 Confidence Bounds Show Marker Calculator Cum Probability Value Statistics M earn 5 2751E 5 Mean Conf 5 954 5 23780E 57 5 3124E 5 S D 22465 Skewness 0 0050336 Eurtosis 3 0424 Reals 100 ne Boos Probability Density b 33488E 5 Cond T all Expectation Result Array Close Manual for PREMChlor Contents e 70 1 Distribution Cost_ source M Pg Em A ete Percentiles Cum Prob 0 001 0 01 13220 93 13478 5 15578 2 17695 1 22660 31587 454994 56112 3 6 3203 6 Frare3 8451 1 2 oo Statistics 34432 Mean Cont 52957 32022 2 35962 5 0 14875 3 169 Mean Skewness 0 82903 Kurtosis Mum Reals 100 4 0 05 3 05 2 05 1 0 05 gt 2 4 0e04 B 0 4 PDF Confidence Bounds Show Marker Calculator Cum Probability Value 2 5 467 76 5 Probability Density Cond T ail Expectation Result Array Distribution Cost plume ith Ba 5 ER LA Cost plume
25. triangular PDF is defined by a minimum value a most likely value and a maximum value In order to run the model with the enhanced aqueous phase decay rate option for bioremediation technology select both bioremediation technology and enhanced aqueous phase decay rate Manual for PREMChlor Contents 46 The model uses the probabilistic value for the enhanced decay rate unless the deterministic value is checked If the enhanced aqueous phase decay rate is selected it overrides the aqueous phase source decay rate otherwise the natural source decay rate 1s applied Unit Cost Unit cost of source remediation technologies cubic meter The unit cost is assumed to have a beta distribution based on the data by McDade et al 2005 Liang 2009 The beta PDF is defined by a mean a standard deviation a minimum value and a maximum value Currently four technologies are functional To compute the remediation cost for a specific technology select the corresponding technology check box The model uses the probabilistic value unless both the specific technology and the corresponding deterministic value are checked For each run only one type of the technology can be selected Transport Parameters source plume function 5 Darcy Velocity 2 a Mean If checked use Deterministic Value Manual for PREMChlor Transport Parameters Porosity Retardation Factor Min Lik
26. 0053591 1 87121505 00445 65521505 0 037103 1897 485 4 9427e 05 0 030872 18512 1094 372282 05 __0 017845 180 55 5 o 001568 17433 o 00082957 E of 00082442 161 02 30363 0 000085029 153 55 of 0000309 14249 01 000021453 EE s of 00001725 12677 00001 38 37 0 65708 2 796 25 176 120 52 nada NS x5 Steps 11 16 for viewing the Total Concentration Steps 11 16 can be used to view the Total Discharge and Total Cancer Risk respectively 17 Click the Total Cost tab from the main interface to show the probability distribution summary screen Manual for PREMChlor Contents e 38 Mi Distribution Cost total jg B GA LA B Cost total in Percentiles Cum Prob 0 001 2 09069 006 2 41198 006 3 14353 006 3 84153 006 58976 006 8 74882 006 1 25861E 007 147558E 007 4 59882E 007 1 91519 007 2 01257 007 C Confidence Bounds Statistics Calculator Mear 9 3545E B5 Cum Probability Value Mean Conf 5 95 8 5 1 0039E 7 41241E 6 Skewness 0 39415 2 411 Reals 100 Probability Density Cond Tail Expectations 1 2837 7 Resul Array This screen shows the distribution statistics including the mean standard deviation skewness and kurtosis and the histogram of the PDF for the total remediation cost Click the Source Cost tab and the Plume Cost ta
27. 14 2 4 I 14 03 1 1 24 1 1 Treatment Zone Treatment Zone Decay Fate 3 2 oa u Fl ii Decay Rate 2 1 Decay Rate 3 1 os it wl n Min Likely If checked use If checked use Deterministic Value 2 4 1 1 2 1 1 Min Likely Max Min Likely x1 400 x2 700 Distance From Source Meters The nine possible plume reaction zones are defined by entering two times and two distances Time Period 1 Defines the time tplume 1 when plume reaction rates change from their initial value which started at 20 Note that this time is unrelated to the time of source remediation Units are years Time Period 2 Defines the time tplume 2 when plume reaction rates change from their previous value which started at t tplume 1 Note that this time is also unrelated to the time of source remediation Units are years The third time period occurs after tplume 2 Defines the first distance zone for reaction rates m X2 Defines the second distance zone for reaction rates m The third zone occurs beyond X2 The time period 1 time period 2 X1 and X2 are deterministic parameters Plume Species First Order Decay Rates A total of 36 first order reaction rates are entered in the model 9 for each species The plume treatment efficiency is measured by reaction rate For each component two of nine reaction zones Zone 1 and Zone 2 i
28. 2003 Parker and Park 2004 Zhu and Sykes 2004 Falta et al 2005a Falta 2008 C xo C where is the flow averaged source concentration corresponding to the initial source mass The exponent determines the shape of the source discharge response to changing source mass Figure 1 When 1 the source mass and source discharge decline exponentially with time Newell and Adamson 2005 and Newell et al 2006 When gt 1 the source 15 never fully depleted and the source discharge 15 always greater than zero When 1 the source is eventually depleted and the source discharge equals zero in the end When 0 5 the source discharge declines linearly with time When 20 the source discharge remains constant until the source is completely depleted Falta et al 2005a Falta 2007 and 2008 C C 0 1 Figure 1 Power function illustration of source mass and source discharge relationship Field laboratory and theoretical evaluations of the source mass source discharge response suggest that may vary between about 0 5 and 2 at real sites Rao and Jawitz 2003 Falta et al 2005a Newell and Adamson 2005 Fure et al 2005 Jawitz et al 2005 McGuire et al 2006 Newell et al 2006 Simulation studies suggest that sites with DNAPL located predominantly in low permeability zones exhibit gt 1 and sites with DNAPL in high permeability zones exhibit 1 Falta et al 2005 a b
29. 21 ____5 9987 19765 _ 80753 0 00027087 034496 _ 55913 _____70 135 0 00020387 08346 _ 52182 1720m _ 6073 000015355 073715 48665 16047 _____52 414 0 00011567 065111 4 5404 1457 ____45 072 87121605 057512 42383 13567 _ 38502 5 56216 05 050802 _ 3 9527 13031 _ 32811 3 5582 12158 ____27 518 37229 05 039643 34415 23548 2 8042e 05 03509 32114 21121 05 ___0 2 9967 80 0012003 __040232 098758 16 417 1 9 27985 13 542 1 1983 05 2 5098 Lg 00103 034496 085988 ___11 081 9 025806 021335 24358 88 00095424 031933 08023 14 67984208 01885 2273 88 00084 029573 0724876 71137 512076 06 016855 24215 3857e 08 014716 Manual for PREMChlor Contents e 37 16 Click on the tab of Probability Histories from the table view to view the probability history table The columns correspond to the statistic of the all realizations including the mean median lower upper bounds 5 95 percentiles and 25 75 percentiles lolx aaa AoE he Tme LowerBound 5 2 Medan 75x 9 lal __ 7 o o o o 9 EE OM 30 3160 3268 2 3397 7 3444 2 3486 8 3514 3 pom 722 2605 2930 3 3360 3 3524 7 3680 5 3783 4 8 _____ 2075 msa 2998 amra 3204 359 3144 52 18532 18055 44475 92 ___18382 25694 91908 ____ 05 198 1317 2886 29829 GOO04zN8 041246 4168 1952 9046 2407 000011557
30. 3 18 3003 8 2859 4 2935 3 24169 393 86 37 233 15 568 14188 14 064 14 044 14 024 14 005 13 885 13 985 13 945 13 926 13 906 13 887 13 857 13 848 13 828 497 5 494 69 431 88 459 09 496 3 4a 57 500 34 2035 159 Percentile Lower Bound Upper Bound ar42 7 3736 4 37301 3243 4 1273 5 347 31 926 12 92217 919 25 315 43 913 62 910 81 908 905 2 902 41 099 62 596 84 834 881 28 888 52 nnn 7h v ri Contents 64 Tutorial 4 Probabilistic Thermal Remediation of PCE Source with Enhanced Plume Degradation 2010 1985 Manual for PREMChlor This tutorial example is based on the previous source remediation case but we will now add an enhanced bioremediation of PCE and TCE in the dissolved plume Here the reductive dechlorination of PCE and will be enhanced for the first 300 m sustained indefinitely from 2010 to 2085 years 25 to 100 This is illustrated graphically using a distance time reaction zone plot that corresponds to the REMChlor plume reaction matrix Natural attenuation Natural attenuation Natural dttenuation 300 Distance from source m In the first 300 m the enhanced biodegradation rates of PCE and TCE from 2010 to 2085 are treated as the stochastic variables In the first 300 m the natural attenuation rates of DCE and VC from 2010 to 2085 are also treated as the stochastic variables The natural attenuation rates of PCE TCE DCE and VC
31. 3 Rem Main Interface Component 4 Component 4 Hem Click the Component 2 Rem tab from the Component 1 Remediation interface to view the plume remediation interface for component 2 TCE the Component 2 Remediation interface Here the reductive dechlorination of TCE only occurs in the first zone and the second zone keeps unchanged as the natural attenuation zone The enhanced biodegradation rate of TCE in the first treatment zone has a triangular distribution with a Min of 0 6 a Likely of 2 4 yr and a of 3 2 Iplume 2 100 Treatment Zane Treatment zone c m Treatment Dimensions Costs Treatment Dimensions Costs gt 5 a T E Decay Rate 2 2 If checked use in Likely Min Likely Deterministic Value 05 24 32 03 12 32 12 tplume Manual for PREMChlor Contents e 66 Click the Component 3_Rem tab from the Component 2 Remediation interface to view the plume remediation interface for component 3 DCE the Component 3 Remediation interface Here DCE keeps the natural attenuation process in both zones The decay rate of DCE in the first zone has a triangular distribution with a Min of 0 1 a Likely of 1 2 yr and a Max of 20 9 tolume 2 100 Treatment Zone Treatment Zone de d Treatment Dimensions Costs Treatment Dimensions Costs 2 5 Decay Rate 1 2 Decay Rate 2 2 E lf checked use k Likely Max Deterministic Pay Min Likel
32. 3074 707 04 12853 635 54 57 _ 10784 57326 516 22 4037 1001 2 464 86 691 96 964 65 1204 1 49 27 12818 418 62 46 76 929 37 1184 5 1336 7 1522 58 37 068 104 44 880 57 27 836 85114 21 327 339 51 355 08 862 49 1148 Figure 7 Probability histories of an output Table view The probability distribution summary includes the distribution statistics such as the mean standard deviation skewness and kurtosis and the PDF for an output The histogram of the PDF is generated by placing the final values of an output from all of the realizations into a discrete number of bins The PDF of an output reflects the overall uncertainty posed by the uncertainties in the input parameters An example of the probability distribution summary for an output is shown in Figure 8 The left table shows the distribution s percentiles below which the distribution statistics are shown The histogram on the right side is the PDF Distribution Cost_total Percentiles 2 05099E 006 2 403527E 006 3 161 94E 006 3 0341 006 6 6767 7 006 6 5217E 006 1 25997E 00 1 47396E 00 1 70121E 00 1 91653E 00 2 01339E 00 Statistics 3 3749E 6 Mean Conf 57957 amp 6901E 6 1 006E 7 4 12 6 0 35133 2 4102 Mum Reals 100 Mean kewvness E urtazis 5e 1 0e 07 z eu 1 0e07 PDF Confidence Bounds Show Marker Calculator Cum Pro
33. 8 Chang 1999 USEPA 2001 Liu et al 2004 Li et al 2007 However only a few models allow running Monte Carlo simulations and stochastic analysis regarding contaminant fate and transport such as MODFLOW2000 within GMS v5 0 http www scisoft gms com and Groundwater Vistas v4 0 http www groundwater models com PREMChlor has been developed by the linkage between the deterministic REMChlor model and the probabilistic simulation package GoldSim http www goldsim com GoldSim is a probabilistic simulation software package for visualizing and conducting dynamic probabilistic simulation to support management and decision making in business engineering and science GoldSim User s Guide 2007 It has a great flexibility to link to other external programs and process models GoldSim has been used in the nuclear industry for conducting performance safety assessment calculations Robinson et al 2003 GoldSim conducts the probabilistic simulation using a Monte Carlo approach The Monte Carlo approach is the common technique for propagating the uncertainty in the input parameters of a system to the predicted results and performance In Monte Carlo simulation the entire system is simulated a large Manual for PREMChlor Contents e 14 Variable Variable z a Variable x Valuelof z Deterministic Model Predicted Value Time or distance Variable y Variable z Variable x b Value of x Value of y Value of z
34. 8 85 Jul 02 2025 BB 284 2 4802 57848 58 229 193 04 267 08 1102 2026 67968 24848 5 7751 57 983 191 4 265 32 Jul 02 2027 B7 653 2 4784 5 7655 57 737 189 78 263 56 Jul 01 2028 B7 338 2474 5 7559 57 492 18816 261 81 Jul 02 2028 67 027 24687 57463 57 248 186 55 260 07 lt gt Manual for PREMChlor Contents e 69 From the chart view the remediation goal red dotted line is close to the 95 th percentile From the table view the 95 percentile concentration at 100m in 2025 is about 193 ug l and the upper bound concentration 2025 is about 267 ug l The remediation effort would meet the goal with about 95 certainty Compared to the previous case the addition of the plume bioremediation along with the original source remediation increased the chance of meeting the remediation goal The new design including the source remediation and the enhanced plume biodegradation appears to be robust A remediation cost analysis is also conducted during the simulation Select Total Cost from the main interface to view the statistics and the distribution of the total remediation cost The mean total cost is 527 500 Select Source Cost from the main interface to view the statistics and the distribution of the source remediation cost The mean total cost is 34 500 Select Plume Cost from the main interface to view the statistics and the distribution of the plume remediation cost The mean total cost is 493 000 Filpistribution Cost
35. Deterministic value Ord oi f 054 0 9 054 Inhalation 0 001 0021 005 0021 X Min Most Ma Manual for PREMChlor Contents e 53 Manual for PREMChlor Two sets of excess cancer risk slope factors are entered for each component in Comoponent interface one for ingestion drinking the water and one for inhalation breathing vapors in the shower bathroom and house Note that direct vapor transport through the vadose zone is not included here The units for the slope factors are risk per mg contaminant per kg body weight per day Oral Cancer Risk Slope Factor Lifetime oral cancer risk slope factor risk per mg contaminant per kg body weight per day The oral cancer risk slope factor is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The oral cancer risk slope factor is a site dependent variable do not use the default values The model use the probabilistic values sampled from PDFs unless the deterministic value is checked Inhalation Cancer Risk Slope Factor Lifetime inhalation cancer risk slope factor risk per mg contaminant per kg body weight per day The inhalation cancer risk slope factor is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The inhalation cancer risk slope factor is a site dependent variable do not use the default values The
36. Initial Concentration g L Initial Mass kg Power Function Exponent 0 005 0 01 0 02 500 1620 3000 141 Min Likely Min Likely Mean pz f checked use 0 01 If checked use 1520 lf checked use 1 5 Deterministic Value Deterministic value Deterministic Value Source Dimensions Source Width m Source Depth m Source Length m 3 10 30 0 5 3 10 3 10 30 Min Likely Max Min Likely Min Likely Max If checked use 10 I checked use zz If checked use 10 Iv 10 Iw 3 Deterministic 2 Deterministic value Deterministic value Initial Concentration Initial source zone concentration flow averaged concentration of dissolved chlorinated solvent leaving the source zone g l The initial concentration is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value This variable is site dependent do not use the default value The model uses the probabilistic value unless the deterministic value is checked Initial Mass Initial source zone contaminant mass M kg The initial mass is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and Manual for PREMChlor Contents 44 maximum value This variable is site dependent do not use the default value The model uses the probabilistic value unless the determinist
37. REMcChlor is the probabilistic version of USEPA REMChlor deterministic model Falta 2008 and U S USEPA 2007 The user is assumed to be familiar with the REMChlor deterministic model before running the probabilistic model 2 PREMChlor is run under GoldSim Player software To open and run the model file with an extension name of gsp the user needs to download and install the GoldSim Player www goldsim com version 9 60 SP4 GoldSim 2009 The GoldSim Player may be downloaded from http www goldsim com Content asp PageID 430 or http www goldsim com downloads software GSP9604 exe Make sure to download the version of 9 60 SP4 GSP9604 exe After downloading GSP9604 exe double click it to install GoldSim Player Once GoldSim Player is installed the model file may be opened in two ways One way 15 to double click the model file and the other 18 to open the GoldSim Player then click the OPEN MODEL option and select a model file 3 Fortran Dynamic Link Library DLL file is also required for the simulation The DLL file links the REMChlor Fortran source code to the probabilistic model In order to run the model properly the probabilistic model file with the extension name of gsp and the DLL file with the extension name of dll need to be in the same directory 4 This program should run on any recent personal computer using the Microsoft Windows XP Vista and Windows 7 operating systems Manual for PREMChlor Contents e 8
38. S 16 SECURITY CLASSIFICATION OF 17 LIMITATION OF 18 NUMBER 19a NAME OF ABSTRACT OF PAGES RESPONSIBLE PERSON a REPORT b ABSTRACT c THIS PAGE Same as 74 unclassified unclassified unclassified Report SAR Standard Form 298 Rev 8 98 Prescribed by ANSI Std Z39 18 This work was supported by the US DoD S Environmental Security Technology Certification Program under research project ER 0704 This project was a joint effort between Clemson University GSI Environmental Inc and Purdue University Contents Overview 6 6 Software Installation and Computer Requirements 8 PREMChlor Description 9 Back 9 Models sue Econ 10 Probabilistic Modelit ADDLIOSCIS 14 PRENIKChlor TInputs and Outputs ovid a R 17 UnitCostand Remediation BEHGIeBGV 20 Berniediation COSE vao Rel EUR S EUR nutu 21 Graphical User Interface 22 Maim 22 CS asia nce eae ne cana een eats 22 Explor mr pct 22 SOUE ZONE DUI 23 bio a UUE 23 Soure Media
39. The source is assumed to behave according to Equation 2 with an exponent of 1 This type of source behavior gives an exponential decay of the source mass and concentration with time Newell et al 1996 Parker and Park 2004 Zhu and Sykes 2004 Newell and Adamson 2005 The release was assumed to have occurred in 1985 and the initial source concentration was 10 mg l leading to an initial source discharge of 6 kg of PCE per year Initially PCE and its daughter products trichloroethylene TCE dichloroethylene DCE and vinyl chloride VC were assumed to undergo natural attenuation The medians of the decay rates from BIOCHLOR database Aziz et al 2000 were assigned to PCE TCE DCE and VC as 1 1 1 2 yr 1 2 and 1 7 yr respectively The compounds were specified a retardation factor of 2 the longitudinal dispersivity is equal to 0 01 times the travel distance the transverse dispersivity is 1 10 of the longitudinal dispersivity and the vertical dispersivity is 1 10 of the transverse dispersivity It is assumed that the compliance plane is located at the 100 meters downstream from the source In the absence of any type of remediation this release would result in a deterministic total concentration around 3600 ug l at the compliance plane in 2010 and 3400 ug I in 2025 due to flushing process ipi xil PS pet p 1995 2005 2015 2025 2035 Time yr Manual for PREMChlor Contents e 56 Suppose
40. Water Resources Monograph 17 Washington DC American Geophysical Union McDade J M T M McGuire and C J Newell 2005 Analysis of DNAPL source depletion costs at 36 field sites Remediation Spring 2005 9 18 McGuire T M McDade and C J Newell 2006 Performance of DNAPL source depletion technologies at 59 chlorinated solvent impacted sites Ground Water Monitoring and Remediation Vol 26 No 1 73 84 National Research Council NRC 2000 Natural Attenuation for Groundwater remediation Washington DC National Research Council National Academy Press Newell C J R K McLeod and J R Gonzales 1996 BIOSCREEN Natural Attenuation Decision Support System User s Manual Version 1 3 USEPA 600 R 96 087 Washington DC U S USEPA National Risk Management Research Laboratory Newell C J and D T Adamson 2005 Planning level source decay models to evaluate impact of source depletion on remediation time frame Remediation Vol 15 No 4 27 47 Newell C J I Cowie T M McGuire and W McNab 2006 Multi year temporal changes in chlorinated solvent concentrations at 23 MNA sites Journal of Environmental Engineering Vol 132 No 6 653 603 Pankow J F S Feenstra J A Cherry and M C Ryan 1996 Dense chlorinated solvents in groundwater Background and history of the problem In Dense Chlorinated Solvents and other DNAPLs in Groundwater eds J F Pankow and J A Cherry 1 46 Waterloo Ontario Canada Wate
41. action rates are time and distance dependent As illustrated in Figure 2 nine reaction zones are used to represent different conditions downgradient from a contaminant source over the life of a plume The first time zone after the release O t t could represent a period of natural attenuation following the contaminant spill The second time zone after the release lt lt could represent a temporary period of active plume remediation i e enhanced reductive dechlorination The final time zone gt could represent long term conditions in the plume after the plume remediation ended another period of natural attenuation Similarly the distance from the source is divided into three zones For x x one set of natural or engineered biogeochemical conditions are present while downstream at x x x another set of conditions could predominate For x x conditions could again revert back to natural background conditions This reaction zone approach provides REMChlor with flexibility to simulate the effect of plume natural attenuation or plume remediation on different contaminant compounds spatially and temporally Falta 2007 and 2008 Each of these space time zones can have a different decay rate for each chemical or compound Natural Natural Natural attenuation attenuation attenuation VI IX Reductive Aerobic Natural dechlorination degradation attenuation II V t Natural Natural Natural attenuati
42. and the source remediation efficiency The model predicts a likely failure of the original design Compared to the previous case the uncertainty of the source remediation efficiency resulted in a lower chance of meeting the remediation goal Manual for PREMChlor Contents e 63 il Time History 1 0e03 1 0 02 1 0 01 Concentration 1 0 00 Upper Bound 95 Percentile 25 Percentile abes entile Date Time Jul 02 2010 Jul 02 2011 Jul 01 2012 Jul 02 2013 Jul 02 2014 Jul 02 2015 Jul 01 2015 Jul 02 2017 Jul 02 2018 Jul 02 2019 Jul 01 2020 Jul 02 2021 Jul 02 2022 Jul 02 2023 Jul 01 2024 Jul 02 2025 Jul 2 2026 Jul 02 2027 Jul 01 2028 Jul 02 2029 4 blean 3562 1 3548 2 3534 4 2953 3 629 13 253 53 230 66 228 31 22717 225 13 225 1 224 07 223 05 222 03 221 02 220 01 219 218 217 216 01 215 fT Manual for PREMChlor Lower Bound 1995 2005 2015 2025 TERRE 5 33152 3281 4 3267 2705 1 442 BB 58 393 35 316 33 886 33 735 33 59 33 645 335 33 555 33 51 33 465 33 42 33 375 33 331 33 286 33 241 43197 Median 3588 2 3575 3 3562 5 29733 608 37 232 92 210 28 208 26 207 46 206 76 206 07 205 37 204 68 204 203 31 202 63 201 95 201 27 200 59 198 48 198 2 95 3705 3 3697 4 2689 5 3138 3 599 96 949 33 524 7 520 55 51755 514 55 211 77 508 9 506 03 50
43. as considered in the source zone Therefore the reported concentration reduction percentages for parent CVOC compound McGurie et al 2006 were used as the source mass removal percentages in PREMChlor The reported statistics of the concentration reduction percentages for parent CVOC compound were used to determine the distribution function for the source removal efficiency Different types of distribution functions available in GoldSim were tested to fit the reported values It was found that the beta distribution fit the reported value best In PREMChlor the mean minimum min and maximum max values of the interpolated beta distribution are the reported median minimum and maximum values McGurie et al 2006 respectively The standard deviation of the beta distribution was adjusted by matching the interpolated PDF with the histogram generated based on the reported values Due to lack of information the enhanced decay rate which is another option to represent the remediation efficiency of enhanced bioremediation 15 assumed to have a triangular distribution The plume treatment methods mainly are enhanced biodegradation PREMChlor can also simulate permeable reactive barriers PRBs Plume PRB treatment can be modeled by assigning a very high first order degradation rate for contaminant in a narrow reaction zone The application of PREMChlor to a plume PRB treatment can be found in Liang 2009 Due to the lack of information the unit cost and
44. ate 1 2 Decay Rate 2 2 ibe If checked use lf checked use Min Likely Deterministic Value Min Likely Deterministic Value o MESES c n CERE i Matural Attenuation ken gt 5 M T tplume 1 Matural Attenuation Matural Attenuation Matural Attenuation Period 1 amp T 400 Distance From Source Meters To go to the other component for the natural decay rates click the Component tab To go to the other component for the remediation degradation rates click the Component _Rem tab To go back to the main interface click the Main Interface tab Manual for PREMChlor Contents 31 Click on the Treatment Dimensions Costs tab and the following Plume Treatment screen will appear source plume function Plume Treatment i ioj x Plume Treatment Natural Attenuation Matural Attenuation Natural Attenuation Treatment zone Treatment one Treatment width m Treatment Width m 50 Treatment Depth m En Treatment Depth m 5 Unit Costi S cubie meter Unit Cast eubie meter If checked use If checked use Min Likely Deterministic Value Mif Likely Deterministic value Attenuation Time Years 1 10 100 10 1 10 100 10 i Natural Attenuation tplume 1 Natural Attenuation Natural Attenuation 1 400 Distance From Source Meters Present Annual O a
45. b from the main interface to view the source remediation cost and plume remediation cost Manual for PREMChlor Contents e 39 18 The concentration mass discharge and cancer risk of each individual component also can be viewed To do this click the Explore Model tab from the main interface and the following screen showing the top level of the model stucture will appear Ed GS Player sourceplume_thist3_dash4c_risk4_cost2la_triangular 101 12 same 7 Model Model main 28 CancerRisk results E Concentration results H S Cost results J Discharge results 5 8 Interfaces Componenti Fg Componenti Rem z Component 1 Component2_Rem E Component3 8 Component3_Rem E Component4 Component4 Rem i Plume_Treatment EE Source Remediation SourceParameters TransportParameters 9 8 Remediation Transport Model 9 08 Observation Location 2 Plume DecayRates 5 8 Run Properties i Ih ElapsedTime 2 SlopeFactor Yield i File View Model Help 128 RS EVE Interfaces Cost results M b Run Made Press F5 to start simulation Scale 100 Fiter OFF 7 To view the concentration of each individual component click the plus sign on the top left side of the Concentration_results container and the following window will appear Fd GS Player REMChlorGoldSim_illustration File View Model Help 59 68 A T
46. bability Probability Density Cond Expectation Result Array 2857E 7 Figure 8 Manual for PREMChlor Probability distribution summary of an output Contents 19 Unit Cost and Remediation Efficiency PREMChlor considers common technologies for DNAPL source removal and dissolved plume treatment Source remediation methods are thermal treatments surfactant cosolvent flooding chemical oxidation reduction and enhanced bioremediation The efficiency of source remediation is represented by the fraction of mass removed In addition efficiency of enhanced source bioremediation has another option as it can alternately be represented by the enhanced decay rate In PREMChlor each remediation technology corresponds to a specific unit cost cost per volume treated and specific remediation efficiency These parameters are treated as uncertain variables represented by the PDFs The distributions and the parameters of unit costs and remediation efficiencies were interpolated based on literature resources McDade et al 2005 presented a comprehensive cost analysis of DNAPL source depletion technologies This study reviewed more than 60 sites and performed the cost analysis for 36 field sites across the United States that had sufficient size cost and performance data to evaluate The unit costs were reported for enhanced bioremediation 11 sites chemical oxidation 13 sites surfactant cosolvent flooding 6 s
47. bals Information Define Monte Carlo options to carry out a probabilistic simulation and specify the sampling method For Stochastic variables 1 Histories to save 100 Run the Following Realization only Use Latin Hypercube Sampling Repeat Sampling Sequences Random Seed Deterministic Simulation Manual for PREMChlor Contents e 59 Click on the Source Zone Parameters tab from the main interface to view the settings of source zone parameters The Initial Mass and Power Function Exponent are set as the stochastic parameters as the check boxes of deterministic values unselected Put the cursor on Initial Mass kg a tool box will appear and show that the initial source mass has a triangular distribution This triangular distribution is defined by a minimum value Min of 500 kg a most likely value Likely of 1620 kg and a maximum value Max of 3000 kg Put the cursor on Power Function Exponent a tool box will appear and show that power function exponent has a log normal distribution This log normal distribution is defined by a geometric mean of 1 and a geometric standard deviation of 1 21 source plume function SourceParameters Initial Concentration g L 0 005 0 01 0 02 Min Likely If checked use 001 Deterministic Value Source Width 3 10 30 Min Likely Max m If checked use Deterministic d Manual for PREMChlor Source Parameters P
48. ced Aquifer Remediation Journal of Contaminant Hydrology Vol 23 No 4 303 327 Domenico P A 1987 An analytical model for multidimensional transport of a decaying contaminant species Journal of Hydrology 91 49 58 Falta R W K Pruess I Javandel and P A Witherspoon 1992 Numerical Modeling of Steam Injection for the Removal of Nonaqueous Phase Liquids from the Subsurface I Numerical Formulation Water Resources Research Vol 28 No 2 433 449 Falta R W P S C Rao and N Basu 2005a Assessing the impacts of partial mass depletion in DNAPL source zones I Analytical modeling of source strength functions and plume response Journal of Contaminant Hydrology Vol 78 No 4 259 280 Falta R W N Basu and P S C Rao 2005b Assessing the impacts of partial mass depletion in DNAPL source zones II Coupling source strength functions and plume evolution Journal of Contaminant Hydrology Vol 79 No 1 2 45 66 Falta R W 2007 REMChlor User s Manual Beta Version 1 0 Falta R W 2008 Methodology for comparing source and plume remediation alternatives Groundwater Vol 46 No 2 272 285 Fure A D J W Jawitz and M D Annable 2005 DNAPL source depletion Linking architecture and flux response Journal of Contaminant Hydrology Vol 85 No 3 4 118 140 GoldSim User s Guide v9 60 2007 GoldSim Technology Group Hope B and M Stock 1998 Guidance for use of probabilistic analysis in human health
49. completed all of the results of REMChlor calculations are stored in GoldSim and assembled into probability distributions and probability statistics Specify PDFs for REMChlor input Sample the PDFs parameters GoldSim GUI Monte Carlo loop Run REMChlor FORTRAN exe Specify Monte Carlo parameters Store REMChlor output from each realization Process output in GoldSim GUI Figure 4 Flow chart of the DLL linkage during the probabilistic simulation PREMChlor can be run in two different modes the probabilistic simulation mode and the deterministic simulation mode Under the probabilistic simulation mode model runs multiple realizations Each realization 15 deterministic and uses a different probabilistic value for a stochastic parameter Under the deterministic simulation mode only one realization is run in which a deterministic value is used for every parameter PREMChlor allows two different types of input information either deterministic or probabilistic values Deterministic values are provided as the inputs to the model when the user knows the specific values the Manual for PREMChlor Contents e 16 model requires When the required information is uncertain the user provides probability distribution parameters such as mean standard deviation etc as the inputs to define the distribution for a stochastic parameter PREMChlor Inputs and Outputs
50. d to be constant i is the average annual inflation rate is the average annual interest rate t is the year and n is the total period of time for plume operation and management In Equation 11 the numerator accounts for the total O amp M cost in current dollar considering inflation and the denominator accounts for the interest rate This formula accounts for the inflation and interest factors at the beginning of the second year Manual for PREMChlor Contents e 21 Graphical User Interface Main Interface Once the model is opened the main interface with the Run Controller will appear This main interface shows the overall layout of the model The left hand side of the main interface has six tabs to direct the user to other interfaces On the right side the user can specify the observation location and view the simulation results The various model interfaces are described in the following The Run Controller allows the user to run monitor pause and step through the simulation The Reset tab on the GoldSim Run Controller clears all of the results and allows the user to edit model settings and start a new simulation The Run Controller also allows the user to jump to any interface To do this click Go tab and select the desired interface For more information about the Run Controller see the GoldSim Player Help file To access the GoldSim Player Help file click the Explore Model tab on the main interface and then click the Help menu o
51. degradation rate for plume treatment are assumed to have triangular distributions More information about unit cost and remediation efficiency can be found in Liang 2009 Manual for PREMChlor Contents 20 Calculation of Remediation Cost Remediation costs of source removal and plume treatment are included in the probabilistic simulation model Remediation cost analysis 1 conducted outside the FORTRAN DLL link The total remediation cost consists of the source remediation cost and the plume remediation cost For source remediation the probabilistic model considers a one time capital cost which is the product of the unit cost of the source remediation and the volume of the treated source zone For plume remediation cost includes a one time capital cost and a total operation amp management O amp M cost in present net value NPV for a certain remediation period The probabilistic model allows two plume remediation zones For each remediation zone the one time capital cost is the product of the unit cost of the plume remediation and the volume of the remediation zone The calculation of the total O amp M cost in NPV is based on the formula in ITRC 2006 TE TotalNPV Y AnnualCostinYeartwithInflation Y AnnualCost 1 i 1 1 1 r Id MES 11 AnnualCost AnnualCost 2 L 7 r I 1 1 1 where AnnualCost is the current annual cost it is assume
52. del This section includes a brief description of the REMChlor analytical model based on the work by Falta et al 2005 b Falta 2007 and Falta 2008 For more detailed description of REMChlor please refer to the REMChlor user s manual Falta 2007 REMChlor model is the technical foundation of the new probabilistic model This transport model fully couples the source remediation to the plume remediation It is not specific to any remediation technology The contaminant source remediation is simulated as a fractional removal of source mass at a future time after the initial release plume remediation 15 modeled by considering time and distance dependent decay rates of parent and daughter compounds in the first order sequential decay chain Falta 2008 The source model is based on a mass balance of the source zone where mass 15 removed by dissolution and advection with some type of additional decay Falta et al 2005a and Falta 2008 DES 0 A M t 1 where Q t is the water flow rate through the source zone due to infiltration or groundwater flow C t is the average contaminant concentration leaving the source zone M t is the contaminant mass in the Manual for PREMChlor Contents e 10 source zone and 4 is the additional decay term to account for chemical or biological destruction of mass in the source zone The source mass 15 linked to the source discharge through a power function Rao et al 2001 Rao and Jawitz
53. e yr Upper Bound ercentile 153 Percentile 25 Percentile 5 centile Lower Bound Wear This probability history shows the statistics of the results such as the mean median lower upper bounds and different percentiles A percentile is the value of a variable below which a certain percent of observations fall The 50 percentile median is the value below which 50 percent of the observations may be found From the top to the bottom the statistics are the upper bound purple circles 95 percentile green triangles 75 percentile red dash line median yellow solid line mean light blue stars 2596 percentile red dash line 596 percentile green triangles and the lower bound purple circles respectively Put the cursor at a percentile curve on the graph a message box will appear and show the exact value for that percentile as the Y value The message box also shows the X value Manual for PREMChlor Contents e 33 12 The probabilistic statistics can be viewed another way Right click the chart window and select Show as Areas and the following screen will appear Ifilrime History 95 Percentile T5 Percentile 25 a Pieds 5 Percentile Lower Bound EU TR EST From the top to the bottom the upper outline of the purple area is the upper bound the upper outline of blue diagonal filled area is the 95 percentile the upper outline of the red diagonal filled area is the 75 percenti
54. ely Max Min Likely If checked use i If checked use 0 3333 2 Deterministic value Deterministic value Darcy Velocity Darcy flux velocity in the flow system V The unretarded chemical velocity pore velocity v is the Darcy velocity divided by the porosity The total flow rate through the source zone Q VYZ 7 Darcy velocity unit is m yr The Darcy velocity is assumed to have a normal distribution The normal PDF is defined by a mean and a standard deviation This variable 15 site dependent do not use the default value The model uses the probabilistic value unless the deterministic value is checked Porosity Effective porosity The porosity is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value This variable is site dependent do not use the default value The model uses the probabilistic value unless the deterministic value is checked Contents e 47 Retardation Factor Retardation factor for all dissolved species must be equal to each other R Typically estimated from organic carbon partition coefficient and soil fraction of organic carbon The retardation factor is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value This variable 15 site dependent do not use the default value The model uses the probabilistic value unless the deter
55. f source mass removed during the remediation This approach is not technology specific and it allows for a realistic and mass conservative assessment of the effects of source remediation on source longevity and discharge The source model serves also as a time dependent mass flux boundary condition to the analytical plume model The plume model considers one dimensional advection retardation and three dimensional dispersion with first order decay of parent compound into daughter products The governing equation for the dissolved concentration of each contaminant compound in the plume is as follows Falta et al 2005b and Falta 2008 arc 9 arc ha Ve ON OV YD 9 Ot Ox Ox Oy OZ Manual for PREMChlor Contents e 12 where C is the dissolved concentration and is the retardation factor amp and are the longitudinal transverse and vertical dispersivities respectively v is the pore velocity and rxn x t is the rate of generation or destruction of the dissolved compound due to biological or chemical reactions that may vary temporally and spatially A streamtube approach is used to decouple the solute advection and reactions from the longitudinal dispersion The one dimensional advective streamtube model is characterized by a constant pore velocity and solute retardation factor Plume reactions are included in this advective streamtube model The entire plume is divided into different zones where the re
56. he molecular weight of component 3 divided by the molecular weight of component 2 If component 2 decays without producing any important daughter products this yield coefficient would be zero The yield 3 from 2 is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked nction Component4 Component 4 Component Mame 4 Yield 4 From 3 0 01 032 1 022 Min Likely Yield 4 From 3 Mass of component 4 created by first order decay of component 3 y4 Typically for reductive dechlorination this would be the molecular weight of component 4 divided by the molecular weight of component 3 If component 3 decays without producing any important daughter products this yield coefficient would be zero The yield 4 from 3 15 assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked Manual for PREMChlor Contents e 50 source plume function Component Time Y ears tplume 2 P tplume 1 P m Plume Reaction Zones Es Component 1 Component Mame 1 Zone 3 Decay Rate 1 3 checked use Decay Rate 12 3 Deterministic Value Decay Hate 3 3 Deterministic Value 03
57. hlor Contents 73 Slob W and Pieters 1998 A probabilistic approach for deriving acceptable human intake limits and human health risks from toxicological studies General framework Risk Analysis Vol 18 787 798 U S Environmental Protection Agency USUSEPA 1989 Risk Assessment Guidance for Superfund Volume I Human Health Evaluation Manual Part A PB90 155581 National Technical Information Service Springfield VA U S Environmental Protection Agency USUSEPA 1997 Cleanup of the Nation s Waste Sites Markets and Technology Trends USEPA 542 R 96 005 U S Environmental Protection Agency U S Government Printing Office Washington DC U S Environmental Protection Agency USUSEPA 2001 Risk assessment guidance for Superfund Volume III Part A Process for conducting probabilistic risk assessment US Environmental Protection Agency Washington DC report USEPA 540 R 02 002 U S Environmental Protection Agency USUSEPA 2004b The DNAPL Remediation Challenge Is There a Case for Source Depletion Kavanaugh M C and P S C Rao editors USEPA 600 R 03 143 National Risk Management Research Laboratory Office of Research and Development U S Environmental Protection Agency Cincinnati Ohio USA Wiedemeier T D H S Rifai C J Newell and J T Wilson 1999 Natural Attenuation of Fuels and Chlorinated Solvents in the Subsurface New York John Wiley and Sons Inc Zhu J and J F Sykes 2004 Simple
58. ic value 1s checked Power Function Exponent Exponent l in source concentration versus mass power function Equation 2 The exponet is assumed to have a log normal distribution The log normal PDF is defined by a geometric mean and a geometric standard deviation This variable is site dependent do not use the default value The suggested value for this variable is ranging from 0 5 to 2 Rao and Jawitz 2003 Falta et al 2005a Newell and Adamson 2005 Fure et al 2005 Jawitz et al 2005 McGuire et al 2006 Newell et al 2006 The model uses the probabilistic value unless the deterministic value is checked Source Width Source zone width The source width is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and maximum value This variable is site dependent do not use the default value The model uses the probabilistic value unless the deterministic value is checked Source Depth Vertical thickness of source zone 2 The source depth is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value This variable is site dependent do not use the default value The model uses the probabilistic value unless the deterministic value is checked Source Length Source zone length X m The source length is assumed to have a triangular distribution The triangular PDF is defi
59. ical dispersivity parameter is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked Number of Stream Tubes Number of streamtubes used to simulate longitudinal dispersion The more tubes used the smoother the solution will look but the longer it will take to compute problem execution time is directly proportional to the number of streamtubes used A solution calculated with only 10 streamtubes will still represent the dispersion Manual for PREMChlor Contents e 48 reasonably well but it will not be smooth solution calculated with 500 streamtubes will be smooth but it will take 50 times longer to compute The number of stream tubes is a deterministic parameter Plume Decay Rates and Yield Coefficients chon Component The plume reaction parameters are entered using a table that mimics the distance time reaction graph Note that in the model time starts at zero when the DNAPL spill occurs at location x 0 The 9 different reaction rates must be entered sUSEPArately for each of the 4 chemical species using the Component tabs Component 1 15 always the ultimate parent compound and it is the component that is released from the DNAPL source zone Component 2 is produced from the decay of component 1 and it decays to produce component 3 Component 3 decays t
60. in other reaction zones keep unchanged and use deterministic values Double click the tutorial file Tutorial4 gsp to open it and then click on the Plume Decay Rates tab from the main interface to view the settings of plume parameters for the first component PCE Click the Treatment Rate tab from the Component 1 interface to view the Component 1 Remediation interface which shows the two treatment zones Here the reductive dechlorination of PCE only occurs in the first zone Zone 1 in Period 2 and the second zone Zone 2 in Period 2 keeps unchanged as the natural attenuation zone The enhanced biodegradation rate of PCE in the first treatment zone has a triangular distribution with a Min of 1 1 a Likely of 2 4 and Max of 4 8 Contents 65 source plume function Componenti Rem Component 1 Remediation Component Mame 1 Attenuation Natural Attenuation Period 3 Natural Attenuation Treatment zone Treatment Zone Treatment D imensians Costs Treatment Dimensions Costs Decay Hate 1 2 Decay Rate 2 2 If checked use If checked use Min Likely Deterministic Value Min Likely Deterministic Value 11 24 48 C 14 os u 24 E 11 Attenuation Period 2 i i amp tplume 1 2 Matural Attenuation Matural Attenuation Attenuation Period 1 1 300 Distance From Source Meters Component 3 Component
61. ironmental Protection Agency USUSEPA REMChlor for Remediation Evaluation Model for Chlorinated solvents model Falta 2008 REMChlor is a significant improvement on existing chlorinated solvent transport models because it can simultaneously account for both source and plume remediation REMChlor includes a source model based on a power function relationship linking the source mass to the source discharge and an analytical plume model based on one dimensional advection with three dimensional dispersion The plume model simulates natural attenuation or plume remediation for parent and daughter compounds in the first order sequential decay chain Plume model also calculates the cancer risks posed by carcinogenic compounds assuming that the contaminated water is used in a house for drinking bathing and other household uses PREMcChlor is developed by linking the analytical model REMChlor to a Monte Carlo modeling package GoldSim via a FORTRAN Dynamic Link Library DLL application In PREMChlor all of the uncertain input parameters treated as stochastic parameters represented by probability density functions PDFs The outputs from PREMChlor are probability distributions and summary statistics of the distributions Cost analysis of common technologies for dense non aqueouse phase liquid DNAPL source removal and dissolved plume treatment are included PREMChlor gives users a single platform where cost source treatment plume management
62. it Cost 4cubie meter If checked use checked use Mean Stdv M Max Deterministic Value Mean Sid Min Deterministic Value In M ThermalMethods 097 0025 0 56 05 usi 50 4185 39239 usi 095 ouf 091 1 pl 095 503 56 375 8632 71937 503 56 C oss 0 05 a 16349 65 2616 66752 1634 Bioremediation oss oof oz 3793 25 262 2949 p 3793 checked use enhanced aqueous phase decay 0 8 0 6 0 55 0 8 for remediation efficiency Percent Removed Fraction of source mass at time f that is removed by source remediation technologies X 0 X lt 1 The percent removed is assumed to have a beta distribution based on the data by McGuire et al 2006 Liang 2009 The beta PDF is defined by a mean a standard deviation a minimum value and a maximum value Currently four technologies are functional To simulate a specific technology select the corresponding technology check box The model uses the probabilistic value unless both the specific technology and the corresponding deterministic value are checked For each run only one type of the technology can be selected Note for enhanced bioremediation technology remediation efficiency can be measured by the enhanced aqueous phase decay rate yr in addition to the percent removed The enhanced aqueous phase decay rate is assumed as a triangular distribution The
63. ites and thermal treatment 6 sites The reported statistics were used to determine the distribution function of the unit cost Different types of distribution functions available in GoldSim were tested to fit the reported values It was found that the beta distribution fit the reported value best In PREMChlor the mean minimum min and maximum max values of the interpolated beta distribution are the reported median minimum and maximum values McDade et al 2005 respectively The standard deviation of the beta distribution was adjusted by matching the interpolated PDF with the histogram generated based on the reported unit costs McGuire et al 2006 presented a performance evaluation of DNAPL source remediation technologies at 59 chlorinated solvent contaminated sites Data were collected and complied from similar sources as in McDade et al 2005 The concentration reduction percentages of parent chlorinated volatile organic compound CVOC compound were reported for enhanced bioremediation 26 sites chemical oxidation 23 sites thermal treatment 6 sites and surfactant cosolvent flooding 4 sites Since the mass reduction removal data were not reported we assumed the value of the exponent of Equation 2 in order to estimate the mass reduction removal from concentration reduction percentage By assuming is equals to one the ratio of mass reduction to concentration reduction is 1 1 In PREMChlor only the parent CVOC compound w
64. le the yellow solid line is the median light blue stars are the mean the bottom outline of the red diagonal filled area is the 25 percentile the bottom outline of blue diagonal filled area is the 5 percentile and the bottom outline of the purple area 15 the lower bound Put the cursor at somewhere on the graph a message box will appear and show the percentile range in which the Y value at this point falls The message box also shows the X value and the Y value of the upper limit of that range Manual for PREMChlor Contents 34 13 Probabilistic output could also be displayed as individual realizations equivalent to a deterministic simulation result To do this right click on the above chart window unselect Probability Histories and the following single realization screen will appear This plot shows the output for a single realization To view the other single realizations change the number in the Realization field Put the cursor at somewhere on the curve a message box will appear and show the realization number X value and the exact corresponding Y value at that point Manual for PREMChlor Contents e 35 14 Probabilistic output could also be displayed as the results for all realizations do this right click on the single realization chart window select Show All Realizations and a legend window will appear Then right click on the legend window and select Hide the following screen showing all realizatio
65. ministic value is checked Scale dependent Dispersivity Parameters If checked use Min Likely Deterministic value Longitudinal 0 001 0 01 0 1 0 01 Transverse 0 0001 0 001 0 01 M 0 001 Vertical le 3 0 0001 0 001 I 0 0001 Longitudinal Dispersivity Parameter Scale dependent longitudinal dispersivity parameter The product of this parameter and the flow distance is the longitudinal dispersivity The longitudinal dispersivity parameter is assumed to have a triangular distribution The triangular PDF 15 defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked Transverse Dispersivity Parameter Scale dependent transverse dispersivity parameter The product of this parameter and the flow distance is the transverse dispersivity This is generally 1 10 or less of the longitudinal dispersivity parameter value The transverse dispersivity parameter is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked Vertical Dispersivity Parameter Scale dependent vertical dispersivity parameter The product of this parameter and the flow distance is the vertical dispersivity This is generally 1 100 or less of the longitudinal dispersivity parameter value The vert
66. model use the probabilistic values sampled from PDFs unless the deterministic value is checked Contents e 54 Tutorials Getting Started In this manual four tutorial examples are included In order to run the tutorial tutorial model files with the extension name of gsp and the DLL file with the extension name of dll need to be under the same directory To open a model file double click a tutorial model file To run a model file click the Run Model tab on the main interface To save a model file click the Explore Model tab on the main interface then click the File menu on the top tool bar and select Save or Save As To reset a model file after a simulation is finished click on Reset from the GoldSim Run Controller to reset the model in order to edit the parameter values and run a new simulation To close a model file after a simulation is finished close the model file from the main interface or click the Explore Model tab on the main interface then click the File menu on the top tool bar and select Exit Manual for PREMChlor Contents e 55 Tutorial 1 Deterministic Thermal Remediation of a PCE Source TE Concentration ugil This hypothetical problem starts from a deterministic setup involving a 1620 kg release of tetrachloroethylene PCE from the source zone with a groundwater darcy velocity of 20m yr and an average porosity of 0 33 The source zone has dimensions of X 10 m Y 10 m and Z 3 m
67. monitored natural attenuation and risk assessment can all be evaluated together and where uncertainty can be incorporated into the site decision making process A license free file containing the user friendly graphical user interface GUI has been generated to make PREMChlor available for use by others Manual for PREMChlor Contents e 6 DISCLAIMER OF LIABILITY With respect to PREMChlor software and documentation neither the United States Government Clemson University GSI Environmental Inc Purdue University nor any of their employees assumes any legal liability or responsibility for the accuracy completeness or usefulness of any information apparatus product or process disclosed Furthermore software and documentation are supplied as is without guarantee or warranty expressed or implied including without limitation any warranty of merchantability or fitness for a specific purpose DISCLAIMER OF ENDORSEMENT Reference herein to any specific commercial products process or service by trade name trademark manufacturer or otherwise does not necessarily constitute or imply its endorsement recommendation or favoring by the United Sates Government The views and opinions of authors expressed herein do not necessarily state or reflect those of the U S Government and shall not be used for advertising or product endorsement purposes Manual for PREMChlor Contents e 7 Software Installation and Computer Requirements 1 P
68. mp M Cost Treatment Period 20 itelume2 tplumaT If checked use Min Likely Deterministic value Inflation Rate 4 Usi EE 1000 100000 1 180000 Interest Rate 6 go back to the previous screen click the Treatment Rate tab To go back to the main interface click the Main Interface tab Similar operations steps 7 8 and 9 for other components Manual for PREMChlor Contents e 32 10 The parameters are set to run the illustration problem Move the mouse over an input field and a simple explanation of the input will pop up To run the simulation click Run Model tab or click the Run tab on the Run Controller The Run Controller will show the simulation is running Run Controller REALIZATION ELAFSED E T HESTEF 00 00 03 F RUMMIMS STATUS 11 After finishing of the run small window with the message of Simulation Complete will show up Click OK and the simulation results are ready to be viewed now From the main interface click Total Concentration tab and the following graphical output of the probability history will appear I Bg 2 e Upper Bound 26 vr Y 3460 65 ugil a i met Concentration ugil FL r VENE TENE H poco edu ERE ni I 10 20 30 40 50 60 70 80 90 100 Tim
69. n Period 2 can be used for plume treatment and other seven zones are used for natural attenuation The user specifies the seven natural attenuation rates for each component in the Component interface sUSEPArate interface for each individual species is pulled up by clicking on the appropriate Component tab The user specifies the two treatment rates in the component Rem interface which is pulled up Manual for PREMChlor Contents e 51 by clicking on Treatment Rate tab sUSEPArate interface for each species is pulled up by clicking the appropriate Component _Rem tab If all nine zones are natural attenuation then use natural attenuation rates for treatment rates All decay rates have units of yr All rates are assumed to have triangular distributions The triangular PDF of each decay rate is defined by a minimum value a most likely value and a maximum value These decay rates are site dependent variables do not use the default values The model uses the probabilistic values sampled from PDFs for the decay rates unless the deterministic values are checked Plume Treatment Dimensions and Costs The plume treatment dimensions and costs interface is pulled up by clicking on the Treatment Dimensions Costs tab from the treatment rate interface Two treatment zones have different lengths widths and depths The length for the first treatment zone is X1 and the length for the second zone is X2 X1 All dimensions have units of m Ge
70. n the top tool bar Simulation Settings The Simulation Settings tab directs the user to the interface where the user can specify the simulation duration time and the Monte Carlo settings Under the Time tab the Basic Time Settings allows the user to choose either an elapsed time simulation or a date time simulation In both cases the number of time steps needs to be specified Steps in the Time Phase Settings portion of the dialog For more information about the Simulation Settings see the Player Help file To access the GoldSim Player Help file click the Explore Model tab on the main interface and then click the Help menu on the top tool bar Explore Model The Explore Model tab directs the user to the model structure The user can explore how the model 15 set up and view the details of the model For more information about the model structure see the work by Liang 2009 Manual for PREMChlor Contents e 22 Source Zone Parameters The Source Zone Parameters tab directs the user to the interface that allows the user to specify source zone parameters such as the initial concentration initial mass power function exponent Gamma and source dimensions Transport Parameters The Transport Parameters tab directs the user to the interface that allows the user to specify transport parameters such as darcy velocity porosity retardation factor and scale dependent dispersivity parameters Source Remediation The Source Remediation
71. ned by a minimum value a most likely value and maximum value This variable is site dependent do not use the default value The model uses the probabilistic value unless the deterministic value is checked Source Remediation source plume function SourceRemediation Source Remediation If checked use StartTime End Time Min Likely Deterministic Value Remediatinon Time yr 30 31 Aqueous Phase Source Decay 1yr 0 0 01 0 1 0 01 Start Time Time when source remediation begins t yr The start time is a deterministic parameter Manual for PREMChlor Contents e 45 End Time Time when source remediation ends 7 yr The end time is a deterministic parameter Aqueous Phase Source Decay First order aqueous phase source decay rate by processes other than dissolution and flushing A This might include biological or chemical reactions in the source zone that destroy source E mass yr The aqueous source decay is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked source plume function 5ourceRemediation Source Remediation Ift checked use StartTime End Time Min Likely Deterministic Value Remediatinan Time yr 3l 31 Aqueous Phase Source Decay 0 0 01 0 1 i 0 01 Percent Removed Un
72. nent 3 concen3 dist and component 4 concen4_dist Double click the element concen1_dist to view the distribution of the component 1 Mil Distribution SourcePlumeRiskDLL concent xb L Lo Dx EE ETC Percentiles Cum Prob 0 001 0 04 0 05 T 8 92506 006 0 25 0 01 85886 0 9958049 i 0 75 2 72090 10 4778 0 95 16 4562 0 99 48 0402 POF 0 999 62 4715 Confidence Bounds Show Marker Statistics _ E Calculator he eam 3 42203 l lum Probability Ww alui Conf 5 9652 2 0341 4 8225 uns 0 0958049 Probabilita D ensitu 2 5501 Y Cond Expectation 6 6964 Result Arar SD 8 3969 Skewness 44913 Kurtosis 27 626 Hum Reals 100 Click a different result element to view the concentration of other single component To go back to the main interface click the Go tab on the Run Controller and select Main Contents e 41 19 view the mass discharge of each individual component click the plus sign on the top left side of the Discharge_results container from the top level of model structure window shown in step 18 and the following window will appear GS Player REMChlorGoldSim_illustration File Model Help SURES 52 Container Path mainDischarge results search Options E Sy Model 8 Model
73. nerally the costs of plume treatment include the capital cost treatment volume multiply by the unit cost and the annual operation amp Management O amp M cost To calculate the net present value NPV of the O amp M cost for a certain treatment period an annual inflation rate and an interest rate are also included in the model source plume function Plume_Jreatment o o A H tplume 2 8 Period 2 Period 3 Period 1 Manual for PREMChlor Plume Treatment Natural Attenuation Natural Attenuation Attenuation Treatment zone Treatment Zone Treatment Width m 3 Treatment Width 50 Treatment Depth rm 5 Treatment Depth m 5 Unit Costi icubic meter Unit Costi Sicubic meter Attenuation If checked use If checked use Max Deterministic value Min Likely Deterministic value of ol 400 Distance From Source Meters Present Annual O amp M Cast Treatment Period If checked use fiplume tplumed 1000 100000 Deterministic Value Inflation Rate 56 Main Interface 126 7 180000 Interest Rate 96 6 Contents 52 Treatment Width Plume treatment zone width m Two plume treatment zones have different widths The treatment widths are deterministic parameters Treatment Depth Vertical thickness of the plume treatment zone m Two plume treatment zones ha
74. ns will appear This graph shows the outputs for all realizations Each curve represents the output of the each single realization Put the cursor at somewhere on the curves a message box will appear and show the realization number X value and the exact corresponding Y value at that point Manual for PREMChlor Contents e 36 15 Click on Table View from the top toolbar and the following tabular output screen will appear The columns correspond to the realizations Time History n ht iol Sr a 0 4 3421 3 3437 9 4 35028 sa se 2853 28 37 30063 _ 12 20423 28288 2853 24318 2787 25825 3362 9 2605 3020 3 1935 2 2710 6 1440 3 2496 1 3007 5 3251 6 3503 2 2049 3246 4 3396 7 3160 3366 3438 1 434 2420 8 2406 2580 8 2180 9 802 5 48 2218 8 2203 3 2381 7 1955 2 600 35 1681 6 2394 18 2034 5 2018 6 2216 7 17521 449 63 1475 1 2219 5 L3 1866 3 1850 1 2054 8 1569 5 337 07 1294 2 2057 8 1712 6 1696 5 1905 1405 3 252 89 1135 9 1908 2 189 86 24 1572 1 1556 2 1766 3 1257 6 387 12 1769 8 28 1443 7 1428 1 1637 3 1124 3 142 61 875 54 16415 1326 2 1311 1 1519 1 1005 6 107 47 758 97 1522 8 12186 12M 1409 889846 14129 30224 5059 S8356 13117 35874 ____37373 ____10504 ____16872 _____400 72 4 094 ____ 10748 __0 04235 13596 82337 21242 92732 0 00042278 ____1 07
75. o produce component 4 Component 4 is assumed to decay into a harmless species Each of these space time zones can have a different decay rate a for each chemical species Natural Natural Natural attenuation attenuation attenuation time 2025 Natural Aerobic degradation attenuation 2005 Natural Natural Natural attenuation attenuation attenuation 400 700 Distance from source Yield Coefficients Component 2 Manual for PREMChlor Component 2 Ift checked use Yield 2 Fram 1 0 79 0 79 I 1 Likely Yield 2 From I Mass of component 2 created by first order decay of component 1 y2 Typically for reductive dechlorination this would be the molecular weight of component 2 divided by the molecular Contents e 49 weight of component 1 If component decays without producing any important daughter products this yield coefficient would be zero The yield 2 from 1 is assumed to have a triangular distribution The triangular PDF is defined by a minimum value a most likely value and a maximum value The model uses the probabilistic value unless the deterministic value is checked nction Component3 Component 3 Component 3 Yield 3 From 2 0 01 074 1 pr OS 0 74 Min Likely Yield 3 From 2 Mass of component 3 created by first order decay of component 2 Typically for reductive dechlorination this would be t
76. on attenuation attenuation D IV VID 0 0 X1 X5 Distance from source Figure 2 Illustration of plume space time zones The analytical solution for these multiple reaction zones is derived using the residence time in each zone to develop the batch reaction solution for that zone For more details about the solutions see Falta 2007 Manual for PREMChlor Contents e 13 and 2008 This plume model considers first order parent daughter decay production reactions for a four component system The batch reaction equations and solutions for four compounds are given by Falta 2007 and 2008 Longitudinal dispersion is accounted for by considering a collection of streamtubes with a normally distributed pore velocity Falta 2008 With a mean pore velocity v a standard deviation of pore velocity and a mean front location x longitudinal dispersivity ax at x t is calculated by Equation 10 Falta 2008 1 o O f lo 2p 2P 10 a Transverse and vertical dispersions are modeled using Domenico s 1987 approximation The longitudinal transverse and vertical dispersivities are made scale dependent by being different linear functions of the mean front location The plume model assumes dispersion occurring in the positive and negative y directions but only in the positive z direction Falta 2008 The solution with 3 D dispersion is given by Falta 2007 and 2008 Cancer risks posed b
77. ower Function Exponent 1 1 21 Mean Stdv Ift checked use E Deterministic value Initial Mass ka 500 1620 3000 Min Likely If checked use 1420 Deterministic value Source Dimensions Source Depth m Source Length 05 3 10 10 30 Min Likely Min Likely Max If checked use If checked use g Deterministic Value Deterministic value After viewing the uncertain source parameters settings go back to the main interface and click the Run Model tab Once the probabilistic simulation is completed select the Total Concentration from the main interface to view the probabilistic result of the concentration 100 meters downgradient from the source The remediation goal red dotted line falls between the 75 percentile and 95 percentile The 75 percentile concentration at 100m in 2025 is about 154 and the upper bound concentration is 324 ug l Given the uncertainties in the initial source mass and the power function exponent the model predicts more than 75 probability of meeting the remediation concentration goal Contents e 60 History o yis Ex re EB Concentration 1995 2005 2015 2025 2035 Upper Bound Hz 959 UU 75 Percentile 25 Percentile E Lower Bound amp H Time History Date Time Mean Lower Bound 25 Median 955 Upper Bound Jul 01
78. rGoldSim illustration File View Model Help Gu zB A BQH Container Path mainCancerHisk results Search Options Es risk Time Histories of Cancer Risk EE riskl z X risk dist Ae we we MW EE risk3 risk3 dist risktot risk risk risk3 risk4 E riskd s risk dist risktot risktot_dist eT Distributions of Cancer Risk canc La concent dist i concen2_dist concen3 B risktot dist riskl dist risk2 dist risk3 dist risk4 dist pL sa 3 Kg Containment Class Views Resul Scale 100 Filter OFF risk The top row elements are the time histories of mass discharges for the toal risktot component 1 risk1 component 2 risk2 component 3 risk3 component 4 risk4 and all four components risk The bottom row elements are the distributions of mass discharges for the toal risktot dist component 1 risk1 dist component 2 risk2 dist component 3 risk3 dist and component 4 risk4 dist Click an element to view the cancer risk for a single component To go back to the main interface click the Go tab on the Run Controller and select Main Manual for PREMChlor Contents e 43 Model Input Variables DNAPL Source Parameters and Dimensions Source Parameters Initial Time of DNAPL Release t 0 source plume function SourceParameters Source Parameters
79. rloo Press Park E and J C Parker 2005 Evaluation of an upscaled model for DNAPL dissolution kinetics in heterogeneous aquifers Advanced Water Resource Vol 28 1280 1291 Parker J C and E Park 2004 Modeling field scale dense nonaqueous phase liquid dissolution kinetics in heterogeneous aquifers Water Resources Research Vol 40 W05109 Pope G A and R C Nelson 1978 A chemical flooding compositional simulator Society of Petroleum Engineers Journal 18 5 339 354 Rao P S C J W Jawitz C G Enfield R Falta M D Annable and A L Wood 2001 Technology Integration for Contaminated Site Remediation Cleanup Goals and Performance Metrics Sheffield UK Ground Water Quality Rao P S C and J W Jawitz 2003 Comment on Steady state mass transfer from single component dense non aqueous phase liquids in uniform flow fields by T C Sale amp McWhorter Water Resources Research Vol 39 No 3 1068 Reddi L N ed 1996 Non aqueous phase liquids NAPLs in subsurface environment Assessment and remediation In Proceedings of the Specialty Conference held in Conjunction with the ASCE National Convention November 12 14 1996 New York American Society of Civil Engineers 852 Robinson B A C Li C K Ho 2003 Performance assessment model development and analysis of radionuclide transport in the unsaturated zone Yucca Mountain Nevada Jorunal of Contaminant Hydrology 62 63 249 268 Manual for PREMC
80. screening models of NAPL dissolution in the subsurface Journal of Contaminant Hydrology Vol 72 No 1 4 pp245 258 Manual for PREMChlor Contents e 74
81. shows the plume decay rates and the cancer risk slope factors of the first component source plume function Componenti E j Component 1 Component ame Zone 3 lf checked use If checked use Deterministic value Deterministic Value Det 11 24 wl 11 os 11 24 11 os 11 24 tpl 2 Treatment Zone Treatment Zone Decay Rate 3 2 KONIE Decay Rate 1 1 Decay Rate 21 Decay Rate 3 1 Min Likely Mas Likely Min Likely 1 400 Distance From Source Meters tplume 1 fe 3m Time Y ears aem c If checked use Cancer Hisk Slope Factor Deterministic value Inhalation 0 001 0 021 0 05 4 0 021 Min Mast Main Interface x checked use erministic 1 1 1 1 1 1 To access the other component click the Component tab To go back to the main interface click the Interface tab Manual for PREMChlor Contents e 30 8 Click on the Treatment Rate tab on the Component 1 interface and the following screen will appear This screen shows the plume decay rates of the first component in two remediation zones source plume function Componenti Rem i a Component 1 Remediation Component Mame 1 Matural Attenuation Natural Attenuation Matural attenuation Period 3 Treatment Zone Treatment zone Treatment DimensionsCosts Treatment Dimensions Costs Decay R
82. sign of the Remediation_Efficiency container double click the Thermx_prob element under the Definition window click the Edit tab and the beta PDF window will appear Contents e 62 Mi Beta PDF E Pg Bs 5 8 lA a Parameters hers mean herms stdv Fill Area Calculator Cum Probability Statistics Mean 0 94 Probability Density 13 2506 Std Deviation 003 Expectation 0 96371 S 1 3441 Kurtosis Mat available Close the Beta PDF window and Definition window and then go back to the main interface by clicking Go from the GoldSim Run Controller and select Main After viewing the three uncertain parameters the initial source mass the power function exponent and the fraction of source mass removal of the thermal remediation click the Run Model tab from the main interface After the probabilistic simulation 1s completed select the Total Concentration from the main interface to view the probabilistic result of the concentration 100 meters downgradient from the source From the chart view the remediation goal red dotted line is close to the median From the table view the median concentration at 100m in 2025 1s about 203 ug l and the upper bound concentration 2025 is about 900 ug l The remediation effort is predicted to meet the goal approximately 5096 of the time given uncertainties in the initial source mass the power function exponent
83. situ biodegradation 15 commonly employed if the natural attenuation as a remedy is not sufficient These enhanced biodegradation processes can include reductive dechlorination aerobic oxidation anaerobic oxidation and aerobic co metabolism Wiedemeier et al 1999 National Research Council NRC 2000 Alvarez and 2006 DNAPL source and plume remediation efforts are capital intensive Partial source removal can cost from several hundred thousand dollars to tens of millions of dollars McDade et al 2005 Due to the lower capital costs plume remediation costs are normally considered to be smaller than those for source remediation Plume remediation would be the most cost effective strategy for sites where the source is almost depleted by natural dissolution or other processes Falta 2008 However at some sites source mass is significant Without source removal the resulting plume longevity would require a long period of time to treat and manage The operating and managing cost of plume remediation systems for such sites can be comparable to the source remediation costs For many sites a cost effective remediation design requires some combination of source and plume remediation It is therefore necessary to simultaneously evaluate the transient effects of source and plume remediation Analytical site modeling tools have played important role in the remediation selection process The widely used screening level models BIOSCREEN Newell et
84. some remediation effort 1s proposed in 2010 and the remediation goal is to reduce the total concentration to less than 200 ug l in 15 years following the remediation year 2025 at the compliance plane This tutorial is set up to simulate a very effective thermal remediation of the source that removes 97 of the source mass in 2010 with a period of 0 2 years Double click the tutorial file Turoriall gsp to open it and then click on the different tabs from left side of the main interface to view the simulation settings and the input parameters Source Remediation Start Time End Time Hermediatinan Time 23 45 2 Aqueous Phase Source Percent Removed If checked use Mean eidw Deterministic Value Thermal Methods 0 54 0 03 TA 1 0 97 This tutorial is run under the deterministic simulation mode To view this click the Simulation Settings tab from the main interface then click the Monte Carlo tab and the simulation option window shows that the Deterministic Simulation mode is selected Under the Time tab the time settings show the simulation duration is set as from 1985 to 2085 with a time step of 1 year Simulation Settings Time Monte Carlo globals Information Define Monte Carlo options to carry out a probabilistic simulation and specify the sampling method For Stochastic variables e Probabilistic Simulation Solve Simulation deterministically using Element Determini
85. stic Values Element Mean Values Specified Quantile Manual for PREMChlor Contents e 57 After viewing the simulation settings click OK to go back the main interface and click the Run Model tab to run the model After the simulation 1s completed select Total Concentration from the main interface to view the deterministic result of the concentration 100 meters downgradient from the source The total concentration in 2025 is about 98 which meets the remediation goal red dotted line Based on this simulation it appears that the remediation should work but it does not include any uncertainty Manual for PREMChlor Contents e 58 Tutorial 2 Thermal Remediation of a PCE Source Considering Uncertainties in Source Parameters This tutorial example is identical to the previous deterministic source remediation case except that we will now make two source parameters initial source mass and power function exponent in Equation 2 uncertain and run a probabilistic simulation of a source remediation The remediation concentration goal remains same Double click the tutorial file Tutorial2 gsp to open it Then click on the Simulation Settings tab from the main interface under the Monte Carlo tab the simulation options show that the Probabilistic Simulation mode is selected and the number of Monte Carlo realizations is 100 Click OK to go back the main interface Simulation Settings Time Monte Carlo Glo
86. tab directs the user to the interface that allows the user to specify source zone remediation parameters for different source depletion technologies Plume Decay Rates The Plume Decay Rates tab directs the user to the interfaces that allow the user to set up all of the chlorinated solvent plume parameters such as the plume decay rates yield coefficients and cancer risk slope factors for all components The user also can specify the plume treatment parameters such as treatment rates treatment zone dimensions and costs More details about plume parameter settings are described in the Basic Operation section Viewing Model Output View Graphical or tabular Output Major model output can be viewed by clicking the result tabs on the main interface The default output for Total Concentration Total Mass Discharge or Total Risk Factor versus time for any location is the chart view The default output for Total Cost Source Cost or Plume Cost is also the chart view The user however can switch the output format between the chart view and table view under the output view window for both cases Manual for PREMChlor Contents e 23 Basic Operation The following simple operation illustrates the most basic functions and capabilities of the graphical user interface for PREMChlor 1 Double click the model sample file Sample gsp to start the application The following screen will appear source plume function Main i c x Decision
87. tic modeling approach used in models and tools summarized above does not reflect these uncertainties It is therefore useful to develop a model that can fully assess the uncertainties in process and system parameters in order to select a robust remediation alternative In this section a new probabilistic remediation model PREMChlor is presented The new PREMChlor model takes into account the uncertainties in all major parameters and allows for quick simulations of different combinations of source and plume remediation scenarios to evaluate remediation alternatives PREMChlor is developed by linking the analytical model REMChlor to a Monte Carlo modeling package GoldSim http www goldsim com via a FORTRAN DLL application In PREMChlor all of the uncertain input parameters are treated as stochastic parameters represented by probability density functions PDFs The outputs from PREMChlor are probability distributions and summary statistics of the distributions Cost analysis of common technologies for DNAPL source removal and dissolved plume treatment are included PREMChlor gives users a single platform where cost source treatment plume management monitored natural attenuation and risk assessment can all be evaluated together and where uncertainty can be incorporated into the site decision making process A license free file containing the user friendly GUI has been generated to make PREMChlor available for use by others REMChlor Analytical Mo
88. tion SourceParameters E Source Parameters Initial Concentration gL Initial Mass kg Power Function Exponent 0 005 0 01 0 02 500 1620 3000 1 141 Min Likely Likely Mean lf checked use 0 01 It checked use 1570 If checked use M Deterministic Value Deterministic Deterministic Source Dimensions Source width Source Depth m rosy 3 1 of 3 Min Likely Max Min Likely Max Min Likely Iv Pu eee a 8 desc Deterministic Source Length m Interface To go back to the main interface click the Main Interface tab To jump to another interface click the Go tab on the Run Controller and select an interface Manual for PREMChlor Contents e 27 3 Click on the Transport Parameters tab and the transport parameters screen will appear source plume function TransportParameters Darcy Velocity mm yr 20 Mean Iz If checked use Deterministic value Transport Parameters Porosity 0 2 0 3333 0 4 Min Likely If checked use 03333 0 3333 M Deterministic value Scale dependent Dispersivity Parameters Min Likely Ma Longitudinal 0 001 0 01 Transverse 0 0001 0 001 0 01 vertical 1e5 oom 000 Main Interface ie Retardation Factor Min Likely If checked use 2 Deterministic Value If checked use Deterministic
89. ture of the transport model will appear as shown below The user may explore the every detail of the model as navigating through the model hierarchy For more details about model structure see Liang 2009 To go back to the main interface anytime click the Go tab on the Run Controller and select Main Ed GS Player sourceplume_thist3_dash4c_risk4_cost2la_triangular File View Model Help al gt Container Path Model_main T ransport_Model v Search Options Model 8 Model main J CancerRisk results J Concentration results 7 Cost results 8 8 Discharge results B TY Interfaces Componenti EE Componenti Rem Component2 P Component2 Rem Es Component3 i Component3_Rem Component4 Component4 Rem Main Plume_Treatment Source_Remediation SourceParameters 2 TransportParameters 9 9 Remediation 5 1 Transport_Model 9 08 Observation Location 9 09 Plume DecayRates 5 8 Run Properties deltt ElapsedTime 9 02 SlopeFactor Yield Run Properties Plume DecayRates AM sci De J gt gt Observation Location SlopeFactor Yield Containment Run Mode Press F5 to start simulation Scale 100 Fiter OFF 7 Manual for PREMChlor Contents e 26 4 Click on the Source Zone Parameters tab and the following souce zone parameters screen will appear ES xl source plume func
90. ulation options show that the Probabilistic Simulation mode is selected and the number of Monte Carlo realizations is 100 Click OK to go back the main interface Click on the Source Zone Parameters tab from the main interface to view the settings of source zone parameters The Initial Mass and Power Function Exponent are set as the stochastic parameters as discussed in the previous case Click on the Source Remediation tab from the main interface to view the settings of source remediation parameters The Thermal Methods is selected and the fraction of source mass removal 15 set as the stochastic parameter as the check box of the deterministic values remains unselected Put the cursor on Percent Removed a tool box will appear and show that the fraction of source mass removal has a beta distribution This beta distribution is defined by a mean Mean of 0 94 a standard deviation Stdv of 0 03 a minimum value Min of 0 56 and a maximum value Max of 1 Source Remediation Start Time End Time Remediatinon Time 23 25 2 Aqueous Phase Source Percent Removed Ift checked use Mean Sidy Min Max Determihistit Value I Thermal Methods 0 94 0 03 0 56 1 0 97 view the actual beta PDF of the Percent Removed parameter go back the main interface and click the Explore Model tab click the plus sign of the Remediation container then click the plus sign of the SourceRem_Parameters container continuously click the plus
91. ve different treatment depths The treatment depths are deterministic parameters Unit Cost Unit cost of plume treatment cubic meter Two plume treatment zones may use different methods and have different unit costs The unit costs of the plume treatment are assumed to have triangular distributions The triangular PDF is defined by a minimum value a most likely value and a maximum value These unit costs are site dependent variables do not use the default values The model use the probabilistic values sampled from PDFs unless the deterministic value is checked Present Annual O amp M Cost Present annual O amp M cost of the plume treatment The present annual O amp M cost is assumed to have a have triangular distributions The triangular PDF is defined by a minimum value a most likely value and a maximum value The present annual O amp M cost a site dependent variable do not use the default values The model use the probabilistic values sampled from PDFs unless the deterministic value is checked Treatment Period Defines the time for plume treatment operation and management starting from tplumel and ending at tplume2 yr The treatment period is a deterministic parameter Inflation Rate Annual inflation rate The inflation rate is a deterministic parameter Interest Rate Annual interest rate The interest rate is a deterministic parameter Cancer Risk Parameters If checked use Cancer Hisk Slope Factor
92. y Max Deterministic Value 01 12 209 oif 12 205 12 tplume 1 Click the Component 4 Rem tab from the Component 3 Remediation interface to view the plume remediation interface for component 4 VC the Component 4 Remediation interface Here VC keeps the natural attenuation process in both zones The decay rate of VC in the first zone has a triangular distribution with a Min of 0 4 yr a Likely of 1 7 and a Max of 12 2 2 4 Treatment Zone Treatment Zone 0 x Treatment Dimensions Costs Treatment Dimensions Costs m E 5 Decay Rate 1 2 Decay Rate 2 2 It checked use Min Likely Deterministic Value 0 4 17 122 1 Je Costs for both source remediation and plume treatment will be computed in this problem The source remediation unit cost is made uncertain by unselecting the deterministic value check box The Source Remediation interface shows that the unit cost of thermal treatment has a beta distribution with a Mean of 115 1 m Stdv of 50 m a Min of 41 85 m and a Max of 392 39 m Percent Removed Unit Cost cubic meter If checked use If checked use Mean Stdw Max Deterministic value Mean 50 Min Deterministic Value Thermal Methods 0 94 TE 0 56 1 0 97 115 1 50 4135 39239 1151 Manual for PREMChlor Contents e 67 view the plume treatment dimensions and costs parameters in any Component
93. y carcinogenic compounds in the plume are calculated assuming that the contaminated water is used in a house for drinking bathing and other household uses Falta 2007 The plume model currently considers the inhalation and ingestion cancer risk from water that is piped into the house from a well but it does not consider vapor transport through the vadose zone The calculation approach follows USEPA s method USEPA 1989 and Falta 2007 Probabilistic Modeling Approach A deterministic modeling approach takes a single value for each parameter and gives a single prediction of the system response Figure 3a Typically these single values selected for different parameters are best case estimates or sometimes worst case estimates resulting in overestimates or underestimates of results In reality however the hydrogeologic geochemical and process parameters used in a model are either variable uncertain or both variable and uncertain The deterministic model does not consider the nature of overall uncertainty in a simulation A widely used approach for incorporating this uncertainty 18 probabilistic modeling e g using the Monte Carlo technique where uncertain parameters are represented by PDFs and the result itself is also represented by a probability distribution Figure 3b The probabilistic modeling approach has been widely used to perform risk assessment in contaminated sites USEPA 1997 Hope and Stock 1998 Slob and Pieters 199

Probabilistic Remediation Evaluation Model for - CLU-IN

Contents

Download Pdf Manuals

Related Search

Related Contents