Common Causes of Remediation Failure and Solution Practices

Transcription

1 Do it Right, Do it once Common Causes of Remediation Failure and Solution Practices Presented by Derek Ingram, P.E., R.G., P.G. Midwest Operations Manager July 17, 2018

2 Look Familiar???? You Needed This: But You Got This: Don t Worry, I Got This! Expectations are often different than the results obtained

3 Outline Characterization Remedial Objectives Data Evaluation Design Considerations Design Engineering Pilot Study Field Implementation Conclusion 3

4 Characterization

5 Data Issues Characterization Data Often Does NOT meet Remedial Design Needs Defining extent and COCs is only one step to understanding the site conditions Other parameters that should be collected for design needs: fraction organic carbon/total organic carbon iron and other metals sulfide/sulfate nitrate/nitrite metals ORP emerging contaminant concerns (PFOS, PFOA, PFAS, 1,4-dioxane, etc.) TPH 5

6 Reporting Issues It is assumed that the highest concentration encountered is the highest concentration present Reporting limits above the applicable action level TPH needed Comingled sites.but COCs specific to one source Data collection methods 20 foot well screens or open hole bedrock Identify a concern but do not define zone of impact Potential for cross contamination and concentration dilution Well cluster with two offset 10-foot screens provide much better data Provides for remedial targeting Soil samples below groundwater contact Impact below groundwater CANNOT just be considered as a groundwater issue Only treating groundwater concentrations does NOT address contaminant mass present 6

7 Remedial Objectives

8 Establish Realistic Remedial Objectives Difficult or cost-prohibitive to achieve Stringent / low numerical standards for soils MCL s in groundwater Remedial goal at every monitoring/confirmation point Action levels near or above background levels Arsenic Lead in SW Missouri PAHs in Chicago 8

9 Establish Realistic Remedial Objectives More reasonable goals Contaminant mass reduction (% reduction) Reduction (%) in groundwater concentrations Visible NAPL reduction / reduce NAPL migration Statistical analysis with a maximum not to exceed (ex. 10x) Point of compliance (property boundary); # of points Utilization of risk assessments to establish site-specific targets Technical impracticality demonstration Show TID through remedial failure Pilot scale to control costs 9

10 Data Evaluation

11 Contaminant of Concerns Issues COCs are typically presented as compounds that exceed the objective RFPs for remedial design/implementation most often provide only the COC concentrations Proper design requires an understanding of all compounds present Treatment is not selective to compounds deemed to be issues through numerical screening High concentrations react differently than low concentrations The reactions create differing site conditions A different approach may be needed to address low concentrations and/or the secondary compounds 11

12 Design Considerations

13 Define Cost-Effective Solutions The Promise of New Technologies/ Simpler Solutions: Simple, Better, and Cost-Effective But are we correctly understanding: Complexity and Practical Limitations? How They Work/Fail? The Cost of Failure and Do-Overs? Don t forget to put the Effective In Cost-Effective Don t Skip Proper Design and Evaluation

14 Failure Mechanism Injection Volume vs. Pore Volume Lesser percent pore volume injected Will primarily treat preferential pathways or limited radius from injection point More dependent upon diffusion and groundwater transport of oxidant Higher percent pore volume injected Greater distribution via advective flow Less dependent upon diffusion and groundwater transport of oxidant Typical Applications Use (for cost savings) Less Volume Less Oxidant Less Success Less Cost EPA Staff paper presented in 2017 suggests that effective porosity may be 30% or less of the total porosity 75% has been assumed for several years

15 Common State of the Practice Remedial design using vendor dosing spreadsheets Usually a minimum dosing/application recommended Good start provides Cost-Effective starting point but unlikely the expected level of success Not accounting for sensitive design parameters (not typically in RI): TOD, SOD, etc. COD, BOD, abiotic reactions, etc. (interferences) Scavengers, distribution, etc. (COC specific)

16 Common State of the Practice (cont) Designs are not site-specific Generic with assumptions Assumptions may not be understood by reviewer Provides an easy out for differing site conditions When additional evaluation is recommended by the vendors. It is often ignored.characterization is completed, clients want to see results It worked at.. Injection is done into monitoring wells Quick results over short period Limited ROI Rebound likely False bias of effectiveness and success Affects usage of trend analysis of the data

17 XDD Case Study: Oxygen Release Compound Mass Loading Superfund site: Multiple source/plume with chlorinated solvents and petroleum hydrocarbons Comparison of oxygen release products for petroleum plume in a lab setting Evaluated three oxygen release compounds plus controls Requested dosing recommendations from each product vendor to hit goals Tested three products at the highest recommended dosage of any product* * Occasionally vendors recommend treatability testing to validate dosage assumptions but most often this is not the case and wasn t the case for these vendors

18 Case Study: Oxygen Release Compound Performance Vendor Design Estimates (objective >90% Reduction with Single Dose) % Contaminant Reduction All Products Failed, Even After 3 Applications at the Maximum Dose Recommendation

19 Contaminant Type Contaminant MnO 4 S 2 O 8 SO 4 Fenton s Ozone Petroleum Hydrocarbon G/E G/E G/E E E Benzene P G G/E E E Phenols G P/G G/E E E 1 Polycyclic Aromatic Hydrocarbons (PAHs) G G E E E MTBE G P/G G/E G G Chlorinated Ethenes (PCE, TCE, DCE, VC) E G E E E Carbon Tetrachloride P P P/G P/G P/G Chlorinated Ethanes (TCA, DCA) Polychlorinated Biphenyl's (PCBs) P P G/E G/E G P P P P G 1 P = Energetics poor G = good (RDX, E = excellent HMX) 1=Perozone E? E??

20 Permanganate MnO 4 - Advantages High stability in subsurface Provides better overall efficiency Allows for diffusion into tight soils & porous rock No gas/heat production - less health & safety issues Applicable over wide ph range Many successful in-situ field applications Disadvantages Lower oxidation potential \ Narrower range of contaminant applicability Does not address benzene Metal impurities in product Potential pore clogging due to precipitates

21 Persulfate S 2 O 8 2- Advantages Can be catalyzed by reduced metals or heat to promotes Sulfate Free Radical (SFR) formation High oxidation potential \ applicable to wide range of organics Disadvantages Reaction kinetics heavily dependent on activation technique May have high non-target demand Possible localized low ph

22 Hydrogen Peroxide H 2 O 2 Advantages High oxidation potential \ applicable to wide range of organics The most studied of the oxidizing compounds for remediation Can be combined with ozone (perozone) Disadvantages Reaction s gas/heat production health & safety hazard Short half-life \ limited travel distances, requires closely spaced injection points Optimal ph between 3 5 Ineffective in alkaline environments

23 Ozone O 3 Advantages High oxidation potential \ applicable to wide range of organics Extraction/Oxidation process Easier to apply than liquid oxidants in vadose zone Generated on-site, allows for continual application Disadvantages Less stable than liquid oxidants resulting in shorter half-life Effective distribution in saturated zone requires array of injection points Confined aquifer usage requires pressure (gas) relief Decomposes to oxygen which can stimulate aerobic bioremediation

24 Half-life of COCs Published ranges of compound half-life s can be used to evaluate longterm databases Variances in half-life are due to site conditions but this can provide for a baseline for remedial potential Example: East cost petroleum site with an active bio-sparging treatment system (10 years) in which quarterly groundwater reports identified effectiveness of the systems operation. Quick peer review of data for system optimization purposes. However, revealed that half-life for benzene were similar or hindered in areas treated to non-treated areas over the system operational period. - Likely potential causes were depleted nutrients (just feed it) or changes in geochemistry (presence of increased oxygen reduced sulfate availability) Rather than further investigate and optimize the system, closure through risk-based evaluation was presented and excepted by the agency. 24

25 Design Engineering

26 Feasibility and Bench Studies Technology selection and evaluation Not Research It s the Design Tool Technical evaluation Provides for expertise in design and implementation of most applicable technologies Bench scale evaluation Treatability laboratory and the ability to test most common technologies Field Critical scale for Accurate Cost and Performance Assessment Implement field pilot to full-scale Determine effectiveness under site conditions

27 Objectives for ISCO Bench Testing Determine the ability and rate/dosage of an oxidant to destroy the target contaminants Determine the oxidant demand of the site soils Determine the by-product formation of the oxidationreduction reactions Analyze potential for metals release Determine catalyst or buffering requirements

28 Total Oxidant Demand Tests Soil demand has been shown to vary considerably between soils Can very <1 g/kg to >20 g/kg Factors affecting SOD Organic matter Reduced metals Minerals Applied oxidant concentration Batch studies assume complete mixing May underestimate surface reactions Doesn t simulate subsurface conditions and discrete chemistry (mixing fronts etc.) Oxidant/activator concentration dependent ¹

29 Pilot Study

30 Design Considerations Duration Dependent on site conditions Reaction kinetics Typically days to weeks Oxidant Loading Sufficient oxidant mass for measurable reduction in COC TOD, contaminant mass, distribution Location Representative site conditions Worst case conditions Based on rates of Migration Oxidant consumption Contaminant destruction

31 Single-Well Tests Push-pull tests Inject known volume of oxidant and conservative tracer Extract and analyze change Compare to control test Advantages Minimal equipment needs Short duration (1 to 3 days) Low cost Use existing well* Estimate of SOD Estimate of COC destruction Low volume of reagent used Disadvantages Provides limited information on full-scale delivery method Generates groundwater that may require disposal or treatment $

32 Dual Well Tests Injection / extraction tests (circulation tests) Inject known volume/mass of oxidant and conservative tracer Extract and analyze Advantages Larger aquifer volume tested Better estimation of SOD Better estimation of COC destruction Better estimate of oxidant distribution Low equipment needs Disadvantages Typically requires installation of injection points/wells May or may not be able to reinject extracted water Permitting for re-injection of extracted water $ $ Longer duration (1 to 2 weeks)

33 Multi-Well Tests Multi-point injection Inject known volume/mass of oxidant Monitor multiple points over time Advantages Applicable to all oxidants Enables better ROI determination Able to better simulate fullscale application Disadvantages High cost ($$$) Requires installation of multiple wells Longer duration Higher oxidant batching/injection equip needs $$$

34 Field Implementation 34

35 Ever Changing Environment Optimization Changing site conditions occur immediately upon initiation of remediation activities Remedial design should understand and account for most of these changes in the early stages Data collection and adjustments should be performed periodically Seasonally variations should be addressed Flow rates affect dosage concentration It is not a race Flow rates may be as low as 0.3 to 0.5 gpm Daylighting can occur through preferential pathways.natural or man-made Control through reduced flow rates, reduced pressures (less that 15 psi), and staggered injection cycles Buffering may be needed to postpone chemical reactions Multiple applications may be required to reach sufficient product volume 35

36 Case Study 1: In Situ New York, NY Petroleum Hydrocarbons Treatment with ISCO One shot deal Logistical Issues Tight schedule: complete in 2 weeks Chemical compatibility with construction materials Space limitations: Working around construction activities and maintain traffic accessibility Treatability Study Tested multiple oxidants Alkaline activated persulfate and loading selected Site closed by NYSDEC 92 to 95 % groundwater concentration reduction > 99 % reduction of BTEX, DRO + GRO on soils BTEX (µg/kg) 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 BTEX on Soils Baseline Post

37 Case Study 2: Oxidant Stability Issue Northeast Superfund Site Catalyzed hydrogen peroxide (CHP) selected by Army Corp. for treatment of chlorobenzenes in soil and groundwater Bench tested CHP and persulfate CHP with stabilization agents failed due to instability Iron activated persulfate was appropriate and cost-effective alternate Side by side pilots at site confirmed CHP failure (<1-foot ROI) and persulfate success Persulfate was applied successfully at pilot and full-scale

38 Case Study 3: Enhanced Bioremediation Chlorinated Solvents in Fractured Rock Laboratory treatability study determined: No food / carbon source No appropriate bacteria No Nutrients ph not ideal Adverse Site Conditions Fractured Bedrock Ensure metals mobilization would not be an issue Prior to treatment hot spot area required pump and treat Full-scale applied using pull-push approach adding treatability deficiencies and dosage determinations two applications over 12 month period Remedy successful: the pump and treat system evaluation permitted shut-down 38

39 Case Study 4: Use of Multiple Technologies Midwest Utility site with chlorinated solvent impact and existing groundwater control system with 5-year operation schedule Use of multiple pilot-scale studies combined to serve as full-scale treatment Chemical oxidation (source area and high concentration groundwater) (permanganate shallow and persulfate deep) Chemical reduction (high concentration soils and downgradient passive barrier Bio-augmentation with short- and long-term food sources (high concentration groundwater and leaching zone from cohesive unit) Addressed a 23-acre plume in less than 2 years Contaminant mass reduced to meet risk-based action levels in less than 2 years Groundwater containment system still active but treating ND influent reaches 6 quarter ROD agreement for shutdown this fall 39

40 Conclusions

41 Trial and Error: Common Technology Failure Mechanisms Technology failures are commonly due to: Not providing enough mass of treatment reagents to react with contaminant mass Not addressing COC mass distribution / characterization (e.g., lenses, NAPL, geology) Not identifying Contact Limitations Not performing site-specific design evaluations/treatability The consequences.. Recontamination Rebound No apparent effect, cost and schedule.

42 Reality Check These are Complex Technologies For Complex Problems May appear easy to use, but a lot can go wrong Recognize the practical limits Avoid unrealistic assumptions What is in it? Any different than the old technology? Why would it work better? Site-Specific Issues: One approach does not fit all (geochemical interactions, etc.) Contact (mass and volume) also critical Correct characterization is important for design and success Old Technologies Still Work New products are not necessarily providing better results Failure mechanisms are often similar..contact, contact, contact, volume, volume, volume

43 If At First You Don t Succeed.Start With The Basics: Investigation, Design, Testing They can t sell it if it doesn t work though right? These technologies can/do work, but need to understand limitations Knowing your source, doing the design, and verifying with testing is as important as the technology selection Many failures occur due to under-scoped applications, nothing to do with the technology scavengers

44 Thank You! Questions? Derek D. Ingram, P.E., R.G., P.G. 44