1 Session 5 LNAPL Remediation Technology Overview Part

1 Session 5 - LNAPL Remediation Technology Overview Part 2: Performance Metrics for Hydraulic Recovery

2 Discussed So Far Initiated remedy suite selection based on objectives/LCSM u Introduced hydraulic LNAPL recovery technologies u Introduction • Saturation reduction objective

3 LNAPL Recoverability LNAPL characterization Develop LCSM IBT-1, 2 Sections 3 &4 Identify LNAPL concerns Section 6 Introduction Identify LNAPL objectives, goals, site/LNAPL condition to screen technologies (Screening Step 1: Table 6 -1) Screen technologies: Geology factors (Screening Step 2: Tables A) Screen technologies: Evaluation factors (Screening Step 3: Tables B) Section 7 Minimum data requirements and critical technology Group (Tables C) Section 8 Establish goals and metrics and implement LNAPL remediation Monitor/assess LNAPL remediation performance Demonstrate goals met You are here

4 You Will Learn About Introduction u Hydraulic recovery as one of the selected suite of remedies • • When to recover and where? When to stop recovery? What’s left post-recovery? What happens post recovery? What can hydraulic recovery systems typically achieve? u Hydraulic recovery: LNAPL skimming, Bioslurping/enhanced fluid recovery, Dual pump liquid extraction, and Multiphase extraction u

5 LNAPL Transmissivity – Again? Introduction u Introduced LNAPL transmissivity (Tn) yesterday, now to discuss how it can be used as a metric Why the Tn focus: u Since Tech Reg, Tn has gained traction as a recoverability metric u States have been approached about its use/measure u Tn metric being used for at least one site with state and the U. S. EPA regulatory oversight

6 Recoverability Metrics Recoverability of Mobile LNAPL u Pre-requisite for hydraulic recovery: mobile LNAPL u Mobile LNAPL may or may not be migrating u Treatment train: includes removal of recoverable fraction hydraulically • What metric(s) to use to evaluate LNAPL recoverability across a site? • What metric(s) to determine recovery complete?

7 LNAPL Recoverability Metrics: General Categories u Leading Metrics – to determine if LNAPL can be recovered and where Recoverability Metrics • In-well thickness, LNAPL skimming test, and LNAPL transmissivity u Lagging Metrics - to determine when recovery should end • In-well thickness, LNAPL transmissivity, asymptotic recovery, decline curve analysis, LNAPL-water recovery ratio (cost)

8 Leading Metrics Learning Objectives: Leading Metrics u Understand leading LNAPL recovery metrics, what is the most reliable metric, and how to use this decision variable at a site Make sense to start recovery?

Leading Metrics 9 LNAPL Thicknesses at Hypothetical Site Monitoring Well LNAPL Thickness (ft) 1 0. 2 2 1. 2 3 8. 2 4 4. 5 No LNAPL 5 2. 5 LNAPL At which well would you Target LNAPL recovery? MW 3? MW 2 MW 3 MW 5 MW 4 MW 1 Maybe

10 In-Well LNAPL Thickness as a Recoverability Metric u Traditional metric: recover LNAPL from areas with the largest equilibrium in-well thicknesses down to a specified minimum Leading Metric • Poor metric: correlates unfavorably with LNAPL • • recoverability Does not account for soil and LNAPL properties, soil heterogeneity, and LNAPL occurrence conditions (unconfined/perched/confined) Default standard Easy to understand Inexpensive/very accessible

11 Leading Metric LNAPL Skimming Test u Continuously remove LNAPL from well(s) over a relatively short period of time (hours/days) u Qualitatively evaluate results and compare results u Relatively costly u No standard/protocol u Data can be analyzed to determine Tn

12 Emerging Recoverability Metric: Tn Leading Metric LNAPL Transmissivity (Tn) u Reliable measure of LNAPL recoverability u Incorporates: • • • Soil and LNAPL properties LNAPL saturation In well thickness Soil heterogeneity LNAPL occurrence (confined, perched, unconfined), etc. u Varies directly with recoverability: the higher the transmissivity, the higher the recoverability. u Compared across soil, LNAPL and aquifer types u Relatively inexpensive to determine

13 Tn as a Leading Metric Tn Leading Metric Recoverability and Recovery Learning Objectives: u Understand Tn is an important emerging metric, the issues with Tn, the new ASTM Standard (E 2856), and how to use Tn to meet saturation reduction recovery objectives LNAPL Transmissivity ?

14 Upcoming Discussion Points Review baildown test data u Review Tn calculated from baildown test data, and its relationship to LNAPL recovery potential u Issues with Tn metric and the ASTM Tn Standard (E 2856) u ITRC LNAPL Group’s suggested Tn threshold range u Show Tn can be used to evaluate recoverability over the LNAPL body to meet saturation reduction objectives Tn Leading Metric u

15 LNAPL Baildown Test: Recovery Time Variability ks ee 2 w 1 year LNAPL Thickness (ft) Tn Leading Metric y 1 da 6 months 2 hours Well 1 Well 2 Well 3 Well 4 U Well 4 C Elapse Time (min) Still recovering, expected to ultimately reach ~30 ft due to confined LNAPL

16 Metrics Comparison: Tn vs. LNAPL Well Thickness Recovery Rate Based on Baildown Test Data Recoverability and Recovery Metric Well No. 1 2 3 4 U 4 C Tech. Reg = 0. 1 -0. 8 ft 2/day Approximate LNAPL 1 GPM - Water Gauged Skimming Enhanced Transmissivity Thickness (ft) (GPD) Recovery (GPD) (ft 2/day) 15 40 115 4 34 2 5. 7 0. 2 30 0. 4 0. 7 0. 01 2. 6 120 800 31 5. 4 120 900 35 Recoverability Incorporates Site Metric/Recoverability Indicator Reliability Variables? Correlation LNAPL Thickness Poor No Not consistent LNAPL Transmissivity More reliable Yes Varies directly Site Variables: Soil and LNAPL properties, LNAPL saturation, in-well thickness, soil heterogeneity, and LNAPL occurrence (confined, perched, unconfined)

17 LNAPL Transmissivity Metric Issues Tn Leading Metric u u What Tn estimation method is appropriate under a particular condition and how many measurements are needed? • Small/large LNAPL thicknesses • Unconfined/perched groundwater conditions • LNAPL thickness stable • Fluctuating LNAPL thicknesses Well construction: screen placement & sand pack effects; and well development ASTM standard E 2856 presents approaches for conducting Tn tests: data collection and analysis Standard addresses Tn concerns http: //www. astm. org/Standards/E 2856. htm

18 Other Considerations With LNAPL Transmissivity Metric u How do you interpret a Tn value? • What Tn threshold value/range above which to start/end hydraulic recovery? Tn Leading Metric • How should Tn be applied to determine when to start/stop hydraulic recovery? u u ITRC Tech Reg- 0. 1 to 0. 8 ft 2/day is lower Tn range for effective hydraulic recovery At such low Tn values, LNAPL recoverability is severely curtailed and saturations change little with continued recovery

19 Tn Leading Metric Case Study: LNAPL Recovery vs. Tn u Former Refinery u Mobile LNAPL and Tn distribution defined (LCSM) u Pilot LNAPL recovery in two high LNAPL thickness areas with different Tn • High/low viscosity location

20 Former Refinery Old LNAPL Release In-Well Thickness LNAPL Extent, Not Migrating Transmissivity High Thickness/Low Transmissivity High Thickness/High Transmissivity Pilot Test Location

21 Former Refinery Pilot Test Results (Gallons/Day) (Test time: 72 hours of pseudo-steady state conditions) Tn Leading Metric LNAPL Dual Pump Liquid Skimming Extraction High Transmissivity (>10 ft 2/day) 40 600 Low Transmissivity (<0. 01 ft 2/day) 0 0 • LNAPL hydraulically recoverable in high Tn area • LNAPL not hydraulically recoverable in low Tn area despite greater than 5 feet in-well static LNAPL thickness

22 Former Refinery Transmissivity Tn Leading Metric Old LNAPL Release Hydraulic LNAPL Recovery Areas

LNAPL Transmissivity versus Gauged LNAPL Thickness u Data from multiple sites u LNAPL thickness for a given impact depends on hydrogeologic scenario, soil type, variability in water table 100 “Scatter” LNAPL TRANSMISSIVITY (FT 2/DAY) 23 10 1 0, 01 0, 0001 0 1 10 GAUGED LNAPL THICKNESS (FT) 100 A. Kirkman

Tn Leading Metric 24 LNAPL Thicknesses and Tn at Hypothetical Site Monitoring Well LNAPL Thickness (ft) Tn (ft 2/d) 1 0. 2 0. 05 2 1. 2 0. 5 3 8. 2 0. 1 No LNAPL 4 4. 5 6. 2 LNAPL 5 2. 0 At which well would you target LNAPL recovery? MW 3? No MW 2 MW 3 MW 5 MW 4 MW 1 MW 4? Yes

25 Tn Use As A Leading Metric Example Leading Metric Application Learning Objectives: u Tn applied to Sharkey’s LNAPL: where to recover?

26 Sharkey’s: Applying Tn - which wells to estimate Tn? May 2005 LNAPL Thick (ft) 20062009 LNAPL Thick (ft) May 2010 LNAPL Thick (ft) Test Which Well(s)? 1 0 0 1 Yes 2 0 0 1. 5 Yes 3 0. 5 0 3 Yes 4 0 0 0 No 5 0 0 0 No 6 0 0 0 No 7 0 0 0 No 8 0 0 0 No 9 0 0 0 No 10 0 No MW MW-9 Application To Sharkey’s MW-10 MW-2 MW-3 S. S. MW-1 Flow MW-5 MW-4 MW-6 N MW-7 Estimated LNAPL Extent MW-8 100’

27 Tn Use As A Leading Metric Example Leading Metric Application Learning Objectives: u Understand how Tn can be used to locate LNAPL recovery systems via case example LNAPL: where to recover?

28 How Tn Metric May Be Used To Determine If LNAPL Is To Be Recovered And Where? Example Leading Metric Application Evaluate Tn distribution across site All wells contain LNAPL Limit of LNAPL body Wells in compliance point network Scale of Compliance

LNAPL Transmissivity Distribution Pre-Recovery (Baildown Tests) Example Leading Metric Application 29 >> Tn threshold < Tn threshold Tn contour Scale of Compliance

30 What To Do Next? Select treatment train based on LCSM/objectives u Pilot test remedy (scale dependent) u Install full scale LNAPL recovery system Example Leading Metric Application u

31 Example Leading Metric Application LNAPL Recovery Scenario 1 >> Tn threshold < Tn threshold Scale of Compliance well Recovery well Zone of hydraulic capture

32 Example Leading Metric Application LNAPL Recovery Scenario 2 Scale of Compliance >> Tn threshold < Tn threshold Compliance well Recovery well Zone of hydraulic capture

33 Example Leading Metric Application Harford, IL Refinery Site u LNAPL transmissivity accepted as the leading/lagging metric u This site is large and involves multiple PRPs and has US EPA and state regulatory oversight PRP – potentially responsible party

34 Harford LNAPL Site Summary Example Leading Metric Application u u Over three million gallons of gasoline, diesel and unrefined products released from buried pipelines and from the surrounding refineries over a 40 year period 211 homes and businesses located over a shallow LNAPL Body Seasonal vapor intrusion into homes and business building structures Protecting the Village of Hartford public drinking water supply

35 Example Leading Metric Application Hartford Indoor Vapors the Major Concern at Site (1972)

36 Tn Use As A Leading Metric Example Leading Metric Application Learning Objectives: u Sharkey’s Revisited LNAPL: where to recover?

37 Application To Sharkey’s MW-8 MW-5 MW-1 MW-3 MW-2 0 MW-9 ft bgs Sharkey’s: Mobile LNAPL Extent May 2010 Silty Sand 10’ 2006 30’ 2010 40’ Silty Sand

38 Sharkey’s: Is LNAPL Potentially Hydraulically Recoverable? Application To Sharkey’s MW-10 MW-3 Flow May 2010 LNAPL Thick (ft) May 2010 Tn (ft 2/day) 1 0 0 1 0. 08 2 0 0 1. 5 0. 05 3 0. 5 0 3 3 4 0 0 0 --- 5 0 0 0 --- 6 0 0 0 --- 7 0 0 0 --- 8 0 0 0 --- 9 0 0 0 --- 10 0 --- N MW-1 S. S. 20062009 LNAPL Thick (ft) MW MW-9 MW-2 May 2005 LNAPL Thick (ft) MW-5 MW-4 MW-6 MW-7 Estimated LNAPL Extent MW-8 100’ LNAPL observed in MWs

39 Knowledge Check Hydraulic Recovery Capability For Sharkeys: u Was the Tn in 2006 likely higher than, same as, or lower than Tn in 2010? Why? u Is the mobile LNAPL likely hydraulically recoverable? u What well(s) is/are the best target for hydraulic recovery and why? u What is a reasonable explanation for why the highest Tn is from the well with the greatest LNAPL thickness? Is this situation the general rule?

40 Hydraulic Recovery Capability Knowledge Check For Sharkey’s: u Was the Tn in 2006 likely higher than, same as, or lower than Tn in 2010? Why? Lower. The water table was higher, the LNAPL was trapped as residual, reducing the LNAPL saturation and decreasing the Tn relative to 2010. u Is the mobile LNAPL likely hydraulically recoverable? Yes, in MW-3 area. Reported Tn for MW-3 is higher than the 0. 1 to 0. 8 ft 2/day range suggested by ITRC for potential hydraulic recoverability.

41 Knowledge Check What well is the best target for hydraulic recovery and why? MW-3 because it has the highest Tn. u What is a reasonable explanation for why the highest Tn is from the well with the greatest LNAPL thickness? Is this situation the general rule? The geology at the site is depicted to be consistent across wells. In general as LNAPL saturation increases, in-well LNAPL thickness increases. Increase LNAPL saturation increases Tn. No, in-well thickness is not a reliable indicator of highest Tn, geologic heterogeneity across wells frustrates the metric. Hydraulic Recovery Capability u

42 So, If Hydraulic Recovery of LNAPL… Hydraulic Recovery Capability Is Necessary: How much good are we likely to do? u We’ll visit some case studies u • Emergency release response • Old spill response u What has experience taught us?

Hydraulic Recovery Capability – Emerg Response 43 LNAPL Recovery: Emergency Spill Response Learning Objectives: u Understand what we mean by emergency LNAPL response, and differences in LNAPL recoverability between new and old releases Any better recovery if the spill just happened?

44 Hydraulic Recovery Capability – Emerg Response Emergency LNAPL Recovery u Catastrophic LNAPL release • Large release in a short time • Atypical Acute or potential acute impact on human health/environment u Response immediate u • Abate and control impact/limit footprint • Recover large LNAPL volume in a short time • Immature LCSM

Hydraulic Recovery Capability – Emerg Response 45 Australian Case Example: Recovery While Spreading (Spill Response) Reminder of earlier discussion: u Refinery tank leak of 739, 000 gallons in September 2004 u Isomerate product • Highly volatile, low viscosity. Key ingredient of gasoline. u u Leak a result of corrosion to the tank floor due to low p. H material ‘Homogeneous’ sand

46 Australian Case Example: Monitoring bores Recovery bores Trench Extent of LNAPL 0 ft 130 ft Z Hydraulic Recovery Capability – Emerg Response Quick Installation of Recovery Wells and Trenches N 260 ft

Australian Case Example: T 611 Recovery – Spill Response Monthly Total 300 800 250 700 600 200 Spill Volume (739, 000 gal) 500 Model Calculated Recoverable Volume (617, 000 gal) 150 100 400 300 200 50 100 0 0 Sep Oct Nov 2004 Dec Jan Feb Month Mar Apr May 2005 Jun Cumulative Total (x 1000 gal) Cumulative Total Monthly Total (x 1000 gal) 47

Hydraulic Recovery Capability – Emerg Response 48 Australian Case Example: Estimation of Spill Volume u Analytical tools developed by API and University of Texas used to initially estimate the spill volume u Calculated spill volume approximately 700, 000 gals u Calculated volume within 5% of official estimate Tank Source 0. 8 m 1. 2 m Product Thickness Contours 0. 5 m

Hydraulic Recovery Capability – Emerg Response 49 Australian Case Example: Calculations of recoverable volume: u u u Volatilization during recovery operations • Concentration & flow rate from vacuum tankers Residual hydrocarbon Saturation: 1% • Based on research at site for light hydrocarbon Biodegradation • Based on research rates at site (10 mg/kg soil/day) 82% recovery (617, 000 of 739, 000 gal) Recoverable Volume 82% Volatilization 9% Residual 9% Biodegradation <1%

50 Australian Case Example “LNAPL Table” Surface 2. 7 in Oct 18: 32 Days LNAPL Spreading 8. 7 in Oct 1: 15 Days Ongoing Recovery Case Ex: Dissipating Gradient 1 ft Sep 24: 8 Days Steeper LNAPL Gradient Flatter LNAPL Gradient

Hydraulic Recovery Capability – Emerg Response 51 Australian Case Example: Context – >82% Recovery Why do you think they got 82% out? u Light LNAPL in “sand box” u Emergency response: quick – still appreciable LNAPL gradient to exploit u High saturations (high Tn) to exploit

52 Hydraulic Recovery Capability – Old Spill So, If Hydraulic Recovery of LNAPL… Is Necessary: How much good are we likely to do? u We’ll visit some case studies u • Emergency release response • Old spill response u What has experience taught us?

53 First, Need to Understand LNAPL Recovery and Tn Trends LNAPL Recovery And Tn Trends Learning Objectives: u Gain LNAPL recovery perspective at old sites, recognizing temporal Tn trend with LNAPL recovery, and LNAPL recoverability and Tn metric endpoint Are things different with old spill sites?

Enhanced Skimming LNAPL Recovery Data, Site 4 LNAPL Recovery And Tn Trends 54 2005 2006 Date 2007 2008 2009 2010

LNAPL Recovery And Tn Trends 55 Recovery Declining Tn Declining Recoverability – Self Limiting Relative scale of prior slide

Recovery Rate (gpd) Transmissivity (ft^2/d) LNAPL Recovery And Tn Trends 56 LNAPL Recoverability and Transmissivity Temporal Trends

57 LNAPL Transmissivity Reduction Well 5, Site 2 10 LNAPL TRANSMISSIVITY 0. 3 FT 2/DAY 0. 8 FT 2/DAY CUMULATIVE RECOVERED LNAPL 100% 1 70% 60% 0, 1 50% 40% 30% 0, 01 ESTIMATED POINT AT WHICH LNAPL TRANSMISSIVITY REACHED THE ENDPOINT FOR ENTIRE CAPTURE ZONE Estimated point at which Tn reached the endpoint for entire capture zone 20% 10% DATE 10 вян 08 вян 06 вян 04 вян 02 вян 00 вян 98 вян 96 вян 94 вян 92 вян в- 90 0% ян ян в- 88 0, 001 CUMULATIVE RECOVERED LNAPL (% Of MOBILE LNAPL) 80% LNAPL TRANSMISSIVITY (FT 2/DAY) LNAPL Recovery And Tn Trends 90%

58 What did you Learn? LNAPL Recovery And Tn Trends u LNAPL transmissivity decreases with increasing LNAPL recovery • Reduced LNAPL saturations • Reduced formation thicknesses Tech Reg: Tn threshold range 0. 1 to 0. 8 ft 2/day u Metric immature; uncertainty exists u Range could be refined with more empirical evidence u

63 Hydraulic Recovery – How Much Can You Get Out? Learning Objectives: Hydraulic Recovery Capability u To understand factors that influence hydraulic recoverability and typical recoverability ranges Bottom line?

64 Fraction of Reduced LNAPL Impact Case Study A (Typical Sites) Percentage of Initial Mobile LNAPL Residual LNAPL Fraction (unrecoverable) Source: Kirkman, AECOM Fraction of LNAPL Beyond Proposed Endpoint Range/was not Recovered Fraction of LNAPL Beyond Proposed Recovery Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range 10 years – Elapsed Time 10 years – Estimated Time Wells 23 -24 Site 1 Site 2 Site 3 Site 4

65 Fraction of Reduced LNAPL Impact Case Study A (Typical Sites) Percentage of Initial Mobile LNAPL Residual LNAPL Fraction (unrecoverable) Source: Kirkman, AECOM Fraction of LNAPL Beyond Proposed Endpoint Range/was not Recovered Fraction of LNAPL Beyond Proposed Recovery Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range 10 years – Elapsed Time 10 years – Estimated Time Mobile LNAPL Residual LNAPL Wells 23 -24 Site 1 Site 2 Site 3 Site 4

66 Fraction of Reduced LNAPL Impact Case Study A (Typical Sites) Percentage of Initial Mobile LNAPL Residual LNAPL Fraction (unrecoverable) Source: Kirkman, AECOM Fraction of LNAPL Beyond Proposed Endpoint Range/was not Recovered Fraction of LNAPL Beyond Proposed Recovery Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range 10 years – Elapsed Time 10 years – Estimated Time Residual LNAPL Fraction (unrecoverable) Fraction ofof LNAPL Beyond Proposed Endpoint Fraction LNAPL Beyond Proposed Recovery Range/was not Recovered Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range Wells 23 -24 Site 1 Site 2 Site 3 Site 4

67 Fraction of Reduced LNAPL Impact Case Study A (Typical Sites) Percentage of Initial Mobile LNAPL Residual LNAPL Fraction (unrecoverable) Source: Kirkman, AECOM Fraction of LNAPL Beyond Proposed Endpoint Range/was not Recovered Fraction of LNAPL Beyond Proposed Recovery Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range 10 years – Elapsed Time 10 years – Estimated Time Wells 23 -24 Site 1 Site 2 Site 3 Site 4

68 Fraction of Reduced LNAPL Impact Case Study A (Typical Sites) Percentage of Initial Mobile LNAPL Residual LNAPL Fraction (unrecoverable) Source: Kirkman, AECOM Fraction of LNAPL Beyond Proposed Endpoint Range/was not Recovered Fraction of LNAPL Beyond Proposed Recovery Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range 10 years – Elapsed Time 10 years – Estimated Time Wells 23 -24 Site 1 Site 2 Site 3 Site 4

69 Fraction of Reduced LNAPL Impact Case Study A (Typical Sites) Percentage of Initial Mobile LNAPL Residual LNAPL Fraction (unrecoverable) Source: Kirkman, AECOM Fraction of LNAPL Beyond Proposed Endpoint Range/was not Recovered Fraction of LNAPL Beyond Proposed Recovery Endpoint/Continued to be Recovered with Significant Effort Fraction of LNAPL Within Proposed Endpoint Range 10 years – Elapsed Time 10 years – Estimated Time Wells 23 -24 Site 1 Site 2 Site 3 Site 4

70 Summary of LNAPL Recovery from Older Spill Hydraulic Recovery Capability u Typical spills are 20 to 40 years old • Low LNAPL gradient • Reduced LNAPL saturations with spreading • More viscous through weathering u Typical LNAPL recovery less than 50% u Small percentage of recoverable LNAPL left with the Tn range of 0. 1 to 0. 8 ft 2/day

71 Hydraulic Recovery Capability Knowledge Check u Can we expect to get 82% via hydraulic recovery? u Do we have the inherent LNAPL gradient at old release sites?

72 Knowledge Check Hydraulic Recovery Capability u Can we expect to get 82% via hydraulic recovery? Maybe at a new release site and extra light NAPL – have the LNAPL gradient and high Tn due to extra light NAPL, high saturation, coarse-grained, low heterogeneity u Do we have the inherent LNAPL gradient at old release sites? No. LNAPL has spread and gradient lost – LNAPL is not migrating – which represents the majority of sites you probably deal with unless you’re with a Spill Response Program.

73 Knowledge Check Example Leading Metric Application Hartford LNAPL Site u Will the end of LNAPL hydraulic recovery likely end source remedy at the Harford Site? Why?

74 Knowledge Check Example Leading Metric Application Hartford LNAPL Site u Will the end of LNAPL hydraulic recovery likely end source treatment at the Harford Site? Why? No. The primary concern is indoor vapors. Hydraulic recovery will reduce LNAPL saturation by bulk removal. Hydraulic recovery not get all the LNAPL mass (typically less than 50% of mass). After hydraulic recovery is complete, composition altering technologies (air sparge, SVE, etc. ) will be needed to abate vapor concerns.

75 Lagging Metrics Learning Objectives: Lagging Metrics u Understand what lagging metrics are what the different potential lagging metrics are So, can we stop recovery yet?

76 Lagging Metrics u How do we determine when to end hydraulic recovery? Lagging Metrics • Lagging Metrics u In-well LNAPL thickness u LNAPL Transmissivity (recovery data) • Effective Tn over the capture zone of system • Typically much larger area than baildown test. • May miss local areas of high LNAPL saturation

Lagging Metrics 77 Lagging or Endpoint Metrics – is hydraulic recovery complete? u LNAPL Transmissivity u Asymptotic recovery u Decline curve analysis u LNAPL-water recovery ratio (cost)

78 Desktop Methods Extrapolate: 4, 000 120 3, 500 100 3, 000 80 Cumulative Recovery Rate 60 2, 000 1, 500 40 1, 000 20 Operating Time (days) 0 220 200 180 160 140 120 100 80 60 40 500 20 0 2, 500 Cumulative LNAPL Recovery (gallons) LNAPL Recovery Rate And Cumulative Recovery 0 LNAPL Recovery Rate (gallons per day) Existing System Performance

79 Performance Metric: LNAPL Recovery Decline Curve Analysis 60 50 40 30 Cumulative LNAPL Recovery (gallons) 4, 000 3, 500 3, 000 2, 500 2, 000 1, 500 0 1, 000 10 y = -0. 0213 x + 79. 007 R 2= 0. 8767 500 20 0 LNAPL Recovery Rate (gallons per day) Decline Curve Analysis

80 Performance Metrics: Decline Curve Analysis And Tn Transmissivity Decline Analysis 0. 4 500000 0. 3 400000 To (m 2/day) 300000 200000 Decline Curve Model Production Data 93 19 94 19 95 19 96 19 97 19 92 19 1997 19 1989 1993 Time (Year) 91 0. 0 0 1985 0. 1 90 100000 0. 2 19 Cumulative Production (gallons) Example Cumulative Production Curve Time (Year)

81 Performance Metric: Water-LNAPL Recovery Ratio (Cost) BBL Water Pumped / Lagging Metrics Year BBL Oil Recovered Remove Studies Adjusted Cost* Cost/Gallon 1992 16, 370/1 not available 1993 41, 310/1 not available 1994 10, 550/1 $835, 000 $7. 13 1995 32, 690/1 $732, 000 $13. 43 1996 107, 400/1 $829, 000 $63. 77 1997 99, 390/1 $893, 000 $50. 84 1998 62, 440/1 $1, 013, 000 $29. 63 1999 11, 081/1 $991, 800 $7. 48 2000 105, 069/1 $928, 000 $79. 01 2001 61, 689/1 $766, 000 $27. 15 2002 88, 858/1 $518, 300 $37. 79

82 Summary of Lagging Metrics With most metrics (decline curve analysis, asymptotic recovery, LNAPL-water recovery ratio) is it’s difficult to quantify, hence define generic endpoints u Empirical data has demonstrated that within Tn range of 0. 1 to 0. 8 ft 2/day, LNAPL mobility is curtailed u Tn : emerging metric; some uncertainty as to the end point range u Other metrics best used as lines of evidence to support a Tn metric end point Lagging Metrics u

83 Application of Lagging Metrics Example Application Of Lagging Metrics Learning Objectives: u Understand how to apply metrics for shutting down an operational hydraulic recovery system How do we make the call?

84 How Are Lagging Metrics Used to Reach a Recovery Complete Decision? LNAPL recovery rate diminishing from an operational system u From recovery data Example Application Of Lagging Metrics u • Tn shows decreasing trend and below lagging metric for extended time • Other lagging metric (decline curve analysis, asymptotic recovery, LNAPL-water recovery ratio) adds to the weight of the evidence u Time to shutdown system and re-evaluate Tn on individual wells

85 LNAPL Recovery Scenario 2 Example Application Of Lagging Metrics LNAPL Transmissivity Trend Tn threshold Tn 1 2 3 4 5 6 7 8 Time >> Tn threshold < Tn threshold Scale of Compliance well Recovery well Zone of hydraulic capture

Example Application Of Lagging Metrics 86 LNAPL Transmissivity Distribution Post-Recovery Baildown Test >> Tn threshold < Tn threshold Compliance well Scale of Compliance

87 What’s Next In This Case? Example Application Of Lagging Metrics u Install system at scale of high Tn area • e. g. LNAPL skimming or vacuum enhanced skimming Evaluate Tn at compliant wells over time (at least at seasonal extremes) u Monitor LNAPL well thickness beyond Tn measurement period (cyclic weather: extreme drought/wetness) (e. g. , drought drops water table beyond range of operating system) u

88 Factors to Consider Limits of Hydraulic Recovery u Will get the recoverable fraction (at best) u Leaves behind the residual fraction • Some mobile LNAPL will be left behind also u Residual fraction typically greater than recovered fraction u LNAPL recovery should not be evaluated in absolute terms, but with respect to what’s left behind (the residual) Key Point: If must get more LNAPL out, excavation or phase change technology are your options!

89 Factors to Consider What About the Residual? u Residual LNAPL is immobile as a separate phase u Vapor-phase and dissolved-phase plumes can emanate from the residual LNAPL u Recovering mobile LNAPL can reduce the longevity of the vapor-phase or dissolved-phase plumes but will not eliminate them – Must go after the LNAPL composition to eliminate them!

90 Post LNAPL Recovery: What’s Next? Depends on the LNAPL recovery objectives u Saturation reduction/LNAPL migration objective Factors to Consider u • Goal met • Monitoring u Dissolved/vapor plume • Goal not met • Transition to other technology: e. g. , SVE, air sparging, groundwater pump and treat, etc u Source Elimination • Goal not met • Transition to other technology: e. g. , SVE, air sparging, etc

91 What Did You Learn? u Two general classes of LNAPL recoverability metrics • Leading: when and where to recover? • Lagging: when is recovery complete? Leading metrics: in-well LNAPL thickness; LNAPL skimming test and Tn u Lagging metrics: in-well LNAPL thickness, Tn from recovery data/baildown test, asymptotic recovery, decline curve analysis, LNAPL/water recovery ratio u Tn is an emerging and good LNAPL recovery metric • Accounts for LNAPL, aquifer properties and LNAPL saturations • Quantifiable Take Aways u continued

92 What Did You Learn? u u Take Aways u u u Tn is not necessarily positively correlated to LNAPL thickness, but correlates positively with LNAPL recovery Issues of standard/protocol relating to Tn – ASTM Tech. Reg: Tn threshold range 0. 1 to 0. 8 ft 2/day Can use Tn to focus the recovery where it counts Quick response to new spill maximize success with hydraulic recovery – LNAPL gradient, high saturations Hydraulic recovery of old spills less effective – no LNAPL gradient, smearing, lower saturations If must get more LNAPL out, excavation or phase change technologies are your options! continued

93 Knowledge Check Is the degree of hydraulic recovery at the Australian isomerate spill likely to be experienced at most LNAPLimpacted sites? Why or why not? Are performance metrics of value? Why? How can they be used?

94 Knowledge Check Is the degree of hydraulic recovery at the Australian isomerate spill likely to be experienced at most LNAPLimpacted sites? Why or why not? No, rarely do all of the above factors align as they did in Australia. Typically, low gradient and LNAPL saturation. Are performance metrics of value? Why? How can they be used? Yes. They establish the upfront criteria by which success is measured – provide a design and operational basis. They can be used as leading or lagging indicators of recoverability