修復氯化溶劑場地的工具和方法.pdf

1、Identifying and Delineating Hidden Sources Within a Larger Chlorinated Solvent PlumeKristin Robrock,Ph.D.,P.E.and Peter Mesard,P.E.,P.G.,C.Hg.ExponentOakland,CAOutline Commingled plume problem at site Methodology Results and suggestions for implementationCommingled Plume Problem TCE,PCE,and other ch

2、lorinated solvents are commonly commingled at contaminated sites and in groundwater at industrial urban areas How to distinguish and delineate where one plume ends and another begins?Particularly difficult when different sites have released the same chemicalIs all this ours?Site A being blamed for a

3、ll TCE in plume Objectives:Is all this TCE from Site A?If not,what are the other sources?Can we delineate the extent of Site A TCE?Complicated site with multiple on-site sources Site operated 1907 present Multiple vapor degreasers Degreasers moved locations over time Multiple different solvents over

4、 timeSite AGroundwater flow complicated by hydraulic divide Groundwater divide due to pumping Divide appeared 1970 On-site pumping wells impacted groundwater flow locallyTCE contours suggest other contributors to the plume TCE contours show hot spots and anomalies inconsistent with the direction of

5、groundwater flow TCE flowing onto Site from upgradient source(s)PCE plume looks very different from TCE PCE plume ends shortly downgradient of Site A Similar short plumes for other solvents,like 1,1,1-TCA Therefore,there must be other sources of TCEMany suspicious industrial sites in the area Site i

6、n an urban industrial area Many historical metal plating facilities,machine shops,dry cleaners,and other industrial sites with the potential to have used and released chlorinated solventsPotential chlorinated solvent users may explain TCE anomalies TCE hotspots located adjacent to documented chlorin

7、ated solvent users and some potential users Some sites have soil data indicating TCE releases(Sites C,I,and K)No soil data for many other sites(Sites D,G,L,M,N,P,R,Q)How to identify other sources to this plume?How to separate Site As TCE plume from others?Where does Site As TCE plume end?Methodology

8、 involves 3 steps Sources:Dai Q,and Chau T.2008.Mass separation and risk assessment of commingled contamination in soil and ground water.GeoEdmonton 2008:61st Canadian Geotechnical Conference and 9th Joint CGS/IAH-CNC Groundwater Conference,September 21-24,2008,Edmonton,Canada.Robrock K,and Mesard P

9、.2018.Distinguishing between multiple dry cleaner sources in a comingled chlorinated solvent plume.Battelle 11thInternational Conference on the Remediation of Chlorinated and Recalcitrant Compounds,Palm Springs,CA,April 2018.U.S.EPA.2002.Ground Water Issue-Calculation and Use of First-Order Rate Con

10、stants for Monitored Natural Attenuation Studies.Washington,DC:U.S.EPA1.Prepare“fingerprints”consisting of“pie-charts”of PCE and its daughter products a.Identify fingerprints that match the site and those that dont2.Plot contaminant concentrations along the centerline of a plume a.Identify data poin

11、ts with fingerprints matching the site and those that dont 3.Fit a first-order decay curve to observed data with fingerprints matching the sitea.Extrapolate the distance to meet the desired concentration1.Fingerprinting2.Centerline plot 3.First-order decay extrapolationStep 1:Molar Pie Chart Fingerp

12、rintsIncrease in size of pie suggests additional releaseIncrease in proportion of parent product suggests additional releasePCETCEcis-1,2-DCEvinyl chlorideSite AFingerprints Site fingerprints show mixed PCE,TCE and DCE Predominantly TCE with some DCE leaving the siteFingerprints suggest other contri

13、butors to the plume Loss of degradation products and increase in TCE downgradient indicate another TCE source PCE source at Site BFingerprints suggest other contributors to the plumeFingerprints consistent with Site AFingerprints NOT consistent with Site A showing releases from different sites Loss

14、of degradation products and increase in TCE downgradient indicate another TCE source PCE source at Site BSite A fingerprints with PCE from Site BSite D Sanborn Step 2:Centerline plotDai Q,and Chau T.2008.Mass separation and risk assessment of commingled contamination in soil and ground water.GeoEdmo

15、nton 2008:61st Canadian Geotechnical Conference and 9th Joint CGS/IAH-CNC Groundwater Conference,September 21-24,2008,Edmonton,Canada.0246810020406080100Concentration along centerline(mg/L)Distance(m)Observed dataModel extrapolationSite 1Site 2Centerline Plot data along or close to the centerline In

16、 this case,within 350 ft based on the horizontal dispersion of the plume Source:Xu,M.and Y.Eckstein.1995.Use of weighted least-squares method in evaluation of the relationship between dispersivity and field scale.“Groundwater 33(6):905-908TCE concentrations along centerline02,0004,0006,0008,00010,00

17、012,00014,00016,00001,0002,0003,0004,0005,000TCE(ug/L)Distance(ft)Samples with Site A fingerprintSamples not matching Site A fingerprint02,0004,0006,0008,00010,00012,00014,00016,00018,00020,000010002000300040005000TCE(g/L)Distance(ft)Samples with Site A fingerprintSamples not matching Site A fingerp

18、rintInterpolated data points from contouringDiscrete measured dataStep 3:First-order decay extrapolation Plume attenuation follows first-order decay Extrapolate hidden plume using first-order decay curve=ln ln()Source:U.S.EPA.2002.Ground Water Issue-Calculation and Use of First-Order Rate Constants

19、for Monitored Natural Attenuation Studies.Washington,DC:U.S.EPAFirst-order decay curve for Site A TCEy=-0.0034x+9.3153R=0.505802468101205001000ln(TCE)Distance(ft)y=-0.0039x+10.373R=0.896802468101205001000ln (TCE)Distance(ft)Interpolated data points from contouringDiscrete data pointsUse decay curve

20、to extrapolate end of TCE plume(2 ppb)from Site A Extrapolate to 2 ppb along centerline to find end of Site A TCE plume Manually connect centerline to the edges of the plume Discrete and continuous extrapolations yielded similar plume lengthsSite A TCE plume delineation for sampling event 10 years l

21、ater Fewer data yielded greater variability between discrete and continuous methodsConclusions and suggestions for implementing Simple,easily implementable forensic method using existing,conventional data Works best with:Consistent sampling locations over the years Sampling locations along the cente

22、rline of the plume Recommend comparing discrete measured data and interpolated data from contouring program for better reliability of resultsThank you!Kristin Robrock,Ph.D.,P.E.EGeologic Models Utilizing Environmental Sequence Stratigraphy:An Essential Tool to More Effectively Remediate Contaminated

23、 Groundwater SitesColin Plank,CPGOctober 16,2024AgendaGEOSYNTEC CONSULTANTS1.The Problem:Subsurface complexity impacts performance and causes uncertainty.2.Addressing The Problem:Types ofgeologic models in use3.Best Practice:Environmental Sequence Stratigraphy(ESS)1.What it is and how it impacts a m

24、odels success.4.Project Examples5.ConclusionsThe Problem:The Subsurface is ComplexComplexity Consists of:Stratigraphic Geometry Reality vs.Interpreted Hydro stratigraphic unit continuityLithologic Heterogeneity Cumulative impacts of seemingly small features Van Etten Creek,Oscoda,MIThe Subsurface is

25、 NOT HomogeneousProlonged remediation time frames:Grainsize and Sorting Controls Hydraulic Conductivity(K)Back-diffusion of contaminant mass from fine-grained storage zones often occursSignificance of HeterogeneityGEOSYNTEC CONSULTANTSModified from Gillham and Cherry,1983,Fig.10Significance of Geome

26、tryImpacts hydraulic connectivity,well performance,and/or amendment efficacyWell 1Well2Well 3Well 7*Well 4Well 5Well 6Shallow Interm.DeepWhy is Well 7 Off?Where Else on My SiteMight I expect This?This is the subsurface!Geologic Model Types In Use:Static Cross Sections and Maps:2D Conceptual/Lithostr

27、atigraphic Representations 3D Visualizations:Predominately Lithostratigraphic&Analytical Interpolations Numerical Models:Derive Geometry and Parameterization from AboveYou Need A Better Geologic Model GEOSYNTEC CONSULTANTSFoundational Geology For A Stronger Geologic ModelUse legacy data and understa

28、nding of depositional systems to build a predictive understanding of site stratigraphyHypothesis testing guides investigationEnvironmental Sequence Stratigraphy(ESS):GEOSYNTEC CONSULTANTSESS MethodologyGEOSYNTEC CONSULTANTSSite Aerial Imagery Provides AnalogueESS Can Improve All Model Types Because

29、It Impacts Foundational Geologic Interpretation GEOSYNTEC CONSULTANTSBefore ESS USCS Codes and No Facies AnalogueFacies-Based Stratigraphic Geometry is Key Predictive Value.After ESS-Lacustrine/Deltaic Facies CorrelationsModern Fan Using ESS Correlations to Improve or Replace Lithologic Interpolatio

30、nGEOSYNTEC CONSULTANTSInterpolation ESS Interpretation Using ESS to Inform Numerical Modeling and Optimize RemediationGEOSYNTEC CONSULTANTSGroundwater Monitoring&Remediation 43,Summer 2023,pages 7992 Using ESS to Inform Numerical Modeling and Optimize RemediationGEOSYNTEC CONSULTANTSGeologic Model:S

31、ite is a sandbox Numerical Model Assumptions:Homogeneous and IsotropicPCE Pump and Treat Remediation System Not Meeting Predicted Performance Goals GEOSYNTEC CONSULTANTSModFlow and PEST modeling predicted totalPlume CaptureinjectionextractionExisting Sections Did Not Effectively Correlate or Communi

32、cate Key FeaturesGEOSYNTEC CONSULTANTSLith.logs indicated some heterogeneityGamma indicated some seemingly insignificant claysA loose shallow,intermediate,and deep zonation being usedGeneral Facies Model and Two Key Site FeaturesGEOSYNTEC CONSULTANTSMarine Maximum Flooding(MFS)ClayEstuarine Incised

33、Valley Fill(heterogeneous mix)Upper ShorefaceSands ESS Analysis Reveals and Communicates Key Features GEOSYNTEC CONSULTANTSMapping Leads to Single Targeted Additional Extraction WellGEOSYNTEC CONSULTANTSEvolution of PCE Plume Mass Post ESS Optimization GEOSYNTEC CONSULTANTSConclusion:A Robust Geolog

34、ic Model is Key to SuccessGEOSYNTEC CONSULTANTSEvery site has a conceptual model in use,whether you deliberately created one or not:“The A-Sand,The B-Sand.”“Shallow,Intermediate,and Deep”“The Site is a Sandbox”None of these are particularly geologicThe question is,“Is your model based on Trial and e

35、rror learning?Inherited site lore?USCS code interpolations?OR Sound stratigraphic hypotheses and the predictive framework that results?250 m10 mThank You!Colin PlankColin.PlankG31,00032,00033,00034,00035,00036,00037,00038,0000204060801001201401/30/20141/30/20151/30/20161/30/20171/30/20181/30/20191/3

36、0/20201/30/20211/30/2022CUMULATIVE RECOVERY(GALLONS)MONTHLY LNAPL RECOVERY (GALLONS)RECOVERY PERIODHistorical LNAPL Recovery2014-2022Monthly LNAPL RecoveryCumulative Recovery1992-2013 33,400 gallonsCarbon Dioxide Efflux to Surface(90%-99%of NSZD Signal)BiodegradationOxygen TransportMethane Oxidation

37、VolatilizationDissolution and Electron Acceptor Depletion(1%to 10%of NSZD Signal)Mobile or Residual LNAPLEbullitionEbullitionDissolved PlumeGroundwater Flow and Electron Acceptor Flux(O,NO,SO4)CO2 FluxEquivalentNSZD Rate CO2 FluxEquivalentNSZD Ratemol m-2 s-1gal acre-1 year-1mol m-2 s-1gal acre-1 ye

38、ar-131-B720.04230.053131-MW030.382410.7242931-MW060.06390.021331-MW100.03180.01731-MW170.281780.2112531-MW22NDNDNDNDDefinitonsmol m-2 s-1-micromoles per square meter per secondCO2-carbon dioxideDTW-depth to waterft-feetgal acre-1 year-1-gallon per acre-yearID-IdentificationLNAPL-Light Nonaqueous Pha

39、se LiquidND-not detectedNSZD-Natural Source Zone Depletion Average NSZD Flux Rate Across Fall and Spring Events(gal year-1)3,585Average NSZD Rate Across the Year(gal acre-1 year-1)99.8121Estimated LNAPL Plume Area(acres)32.47Estimated Resulting NSZD Flux Rate(gal year-1)3,2413,929Well IDNSZDFall 202

40、2Spring 2023The former fueling facilities at the site were installed in 1956 over approximately 30 acres and comprised of multiple 8,000-and 10,000-to 20,000-barrel aboveground storage tanks(ASTs)associated with fuel stands and hydrants,filter separators and pump houses.The larger ASTs stored JP-4 u

41、ntil 1993 and then stored JP-8;the 8,000-barrel ASTs stored JP-7.The JP-4 fuel line was replaced with the JP-8 line in 1993.Most of the fueling facilities at the site were decommissioned beginning in December 2015 and replaced with new fueling facilities.Both the JP-4 and JP-8 pipelines were abandon

42、ed in place and sealed with grout slurry.NSZD is a three-part process whereby compounds that comprise light non-aqueous phase liquid(LNAPL)are lost from the subsurface due to naturally occurring processes of dissolution,volatilization,and biodegradation.Carbon dioxide(CO2)is the byproduct of LNAPL b

43、iodegradation and/or aerobic degradation of methane formed from LNAPL biodegradation.Subsurface generation of CO2 above background levels is direct evidence of biodegradation and the verification and quantification of the rate of degradation is the focus of this study.Research has shown that the bio

44、degradation capacity of dissolved-phase petroleum hydrocarbons in the aqueous phase is a small fraction of what potentially can be degraded via methanogenesis of LNAPL.Going forward,consideration should be given to incorporate the LNAPL biodegradation rate and alternatives for enhanced biodegradatio

45、n as part of the final Site remedy.In this regard,the following actions are recommended:n Implement periodic CO2 flux sampling to better understand changing flux rates with time,seasonality,and spatial variance of LNAPL degradation.n The next NSZD evwent should consider using fossil fuel traps at 12

46、 locations within LNAPL extent and one location outside the LNAPL extent for background measurement to better understand the spatial variability in the CO2 flux rate.n The fossil fuel traps should be distributed throughout the footprint of the LNAPL plume with consideration given to the depth to gro

47、undwater,thickness of weathered bedrock and the nature and permeability of surface soils.n Nested wells should be installed and sampled to profile vertical changes in the CO2 concentrations and evaluate dispersion in the deeper parts of the plume.Two locations with high potential for elevated CO2 fl

48、ux were selected(31-MW17 and 31-MW06),two locations with medium potential(31-MW03 and 31-MW10),one location with low potential(31-B72),and one location outside of the LNAPL plume closest to 31-MW22 to Higher methane values,in dissolved-phase and well-head vapor,as well as greater LNAPL thickness wer

49、e anticipated to correlate to higher NSZD rates;however,the survey results do not demonstrate a strong correlation to these relationships.For example,well 31-MW06 had a relatively high dissolved methane value,high well-head vapor methane values,and consistently high LNAPL thickness.However,the fall

50、and spring NSZD flux rates for 31-MW06 were not high compared to those from other locations.One factor that may have contributed to this disparity is the depth to fluids.At locations with the highest calculated NSZD flux rates(wells 31-MW03 and 31-MW17),the depth to LNAPL and water were much less th

51、an in the other wells,and thus the CO2 flux travel path is shorter through the overlying weathered bedrock and alluvium.n Verify LNAPL biodegradationn Assess the LNAPL plume biodegradation raten Update LNAPL conceptual site modeln Evaluation of seasonal changes in the LNAPL biodegradation raten Appr

52、oximate the annual LNAPL biodegradation rate at the Site to compare to mechanical removal methods.The LNAPL plume encompasses approximately 32.5 acres with of the associated dissolved-benzene and MTBE plumes exceeding the maximum plume length allowed by the State of California Low-Threat Underground

53、 Storage Tank Case Closure Policy.LNAPL distribution and transport at the Site are functions of the complex fractured bedrock environment and low LNAPL transmissivity of the formation.In 2012,the results of LNAPL transmissivity testing completed at several wells were an order of magnitude lower than

54、 the range typically considered practicable for mechanical recoverability using skimming pumps.The geology below the site is a complex fractured bedrock environment.The distinction between the weathered and competent bedrock is based on fracture density and weathering observed in the fractures from

55、rock cores and from surface-and downhole-geophysical surveys.Geophysical surveys identified local faulting and a predominant fracture orientation of northeast-southwest(NE-SW)and northwest-southeast(NW-SE).Fractures are in some cases highly weathered and dip at a high angle and can be nearly vertica

56、l.Some boring logs have also identified nearly horizontal saprolite zones indicative of low-angle faulting or exfoliation surfaces common in granitic bedrock environments.Several factors were considered when determining where to place the five traps within the LNAPL plume:n LNAPL presence and thickn

57、ess in monitoring wells,assuming that higher LNAPL thickness would have higher CO2 flux.n Dissolved methane,assuming that higher CO2 flux rates would be associated with areas of higher dissolved-phase methane concentrations.n Methane vapor concentrations at the wellheads,measured during a well-head

58、vapor survey conducted prior to trap deployment,using a LANDTEC GEM5000 Gas Analyzer was used to monitor wells within the extent of LNAPL at the Site.serve as a background location.To investigate seasonal variation,the same locations were used for the November 2022(fall)and March/April 2023(spring)e

59、vents with a goal for each period of deploying the traps for a period of 14 days.GROUP 1REFERENCES:1.E-Flux,LLC.2021.Proposal for Estimation of Natural Source Zone Depletion(NSZD)Rates Measurement of CO2 Fluxes with Fossil Fuel,California.11 November.2.Federal Remediation Technologies Roundtable:Tec

60、hnology Screening Matrix Natural Source Zone Depletion.2023.3.Interstate Technology Regulatory Council(ITRC).2018.LNAPL-3:LNAPL Site Management:LCSM Evolution,Decision Process,and Remedial Technologies.Washington,D.C.:Interstate Technology&Regulatory Council.LNAPL Update Team.March.Battelle 2024 Conference Poster_final_outlines.indd 1Battelle 2024 Conference Poster_final_outlines.indd 15/23/2024 3:30:22 PM5/23/2024 3:30:22 PM