Results from the LLNL Gasoline Spill Cleanup

Transcription

1 Results from the LLNL Gasoline Spill Cleanup A worked example of thermal remediation and site closure for NAPL below the water table. Roger Aines LLNL A Collaborative Effort By LLNL and UC Berkeley, Funded by DOE EM 50 1

2 The Gas Pad cleanup provides examples of the major benefits of thermal methods:!increased volatility of contaminants!rapid mass transfer!rapid diffusion and evaporation!boiling of formation!lower viscosity of water and contaminants!faster chemical reactions 2

3 Dynamic Underground Stripping: Steam & Electric Heat, Vacuum Extraction, Monitoring & Control 3

4 The LLNL Gasoline Spill > 140 ft depth Water table at 100 ft Active shipping and receiving yard Gasoline (auto and airplane) with DCE and DCA 7000 gallons removed in one year of operation Steam system mated to existing pump-and-treat with vacuum extraction Full report at: index.html 4

5 Both the source and surrounding plume were removed: only the source was targeted Treated area Benzene : before treatment After treatment Treated area Thermal remediation at LLNL removed NAPL source region from up to 30 ft below the water table, allowing rapid elimination of surrounding plume. 5

6 Dynamic Underground Stripping vs Conventional Recovery Methods LLNL Gasoline Spill Site Conventional Vacuum Recovery Dynamic Stripping 10, Total gallons per day removed in consecutive operations gpd Cumulative Vapor Wat er Daily Rate 2 gpd /88 7/89 7/90 7/91 2/93 1/94 Date Dynamic Underground Stripping removed vadose zone contamination at ~ 15 times the rate of conventional methods, and groundwater contamination at greater than 60 times the conventional rate. 0 6

7 Contaminant Was Herded To The Spill Center, and Rapidly Removed From The Vadose Zone LLNL Gasoline Spill Site Before and After Experimental Dynamic Underground Stripping Treatment ( ) 7

8 Heat moves readily - you don t have to place it carefully Heated volumes scale by tens of meters; pinpoint location of contamination not required. Even the most impermeable locations can be heated and treated by thermal conduction. Steam tends to trace out the permeable pathways: electricity tends to focus on least permeable material. Primary removal mechanism for VOCs is vaporization: vapor is readily collected and removed. Thermal methods do not require you to spend your entire budget precisely locating the problem - most vendors adjust coverage during system installation. 8

9 Steam moves rapidly into permeable zones; conduction enlarges the heated volume over time 9

10 Is there a best way to add heat NO! It s a lot like drilling; site and vendor specifics can make more difference than technique. Energy flux is important: one yard 3 of soil requires ~100 KW-hour to reach 100 C - whether you use steam, electricity, microwaves or hot air. 0 Steam tends to dominate for deep applications. 0 Electricity has been more widely used for shallow sites. 0 Hot air and hot water carry less heat, work slowly. 10

11 Fundamental requirements for effective thermal remediation Enough heat: don t skimp here 0 The goal is to reach boiling in all contaminated areas 0 Low-heat methods help, but fail to realize full potential Good process monitoring: protects client investment 0 Heating flux (power input) in each well 0 Heated areas 0 Extraction temperatures and contaminant load Good engineering practice: don t try this at home 0 Temperature-compatible materials 0 Large treatment systems to catch all that contaminant! 0 Installers and operators familiar with safety issues 11

12 p era t ure ( C) Recovery increased when boiling temperatures reached the extraction wells, zoomed after heat soak. Patience pays. Temperature and Extraction Rates - First Pass Days (1st Pass) Temperature and Extraction Rate Summary - Second Pass Days (2nd Pass) Temperature (C) Calculated Boiling Point Pumped Water Temperature Liters Gasoline Per Hour vapor Temperature At Recovery Wells Well 808 Vapor Temperature Condensation Rate (Water + Gasoline) Gasoline Recovery condensate 0 2/3/93 2/13/93 2/23/93 3/5/93 3/15/93 Date Liters/Hour Total Condensate 0 T em 100 Boiling Point Liters Gasoline Per Hour Pumped Water Temperature condensate vapor Well 808 Vapor Temperature Gasoline Recovery Calculated Temperature At Recovery Wells Condensation Rate Rate 60 (Water + Gasoline) 0 5/24/93 6/3/93 6/13/93 6/23/93 7/3/93 7/13/93 Date Liters/Hour Total Condensate 0.00 Cyclic Depressurization 12

13 What about mobilization Mobilizing contaminants is the purpose of thermal enhancement. Just like pump-and-treat, hydraulic control is required. 0 Concentrations always increase due to heating, even outside target area in warm water or air. 0 Vapor can spread if vacuum control is lost. 0 No instances of NAPL spreading. 13

14 North B' Ground surface I-I' Ground surface Ground surface South B Mid-process drilling showed that contaminant was moved toward the center for removal Initial Groundwater Removal rate (gallons per day) st pass continuous steam following electric pre-heat Removal Rates 2nd pass huff and puff Feb Mar May-93 1-Jul Aug-93 9-Oct Nov Jan-94 Date Spill Estimate Polishing phase (ARV) residual & electrical heat Cumulative gallons removed Final Groundwater July -95 Depth, ft GEW-710 SVB-GP-001 GSB-1 GSW-15 C403-1 GSW-16 SVB-GP-013 GSB-2 GSB-4 HW-GP-003 GSB-801 TEP-GP-106 (proj. 7') HW-GP-105 (proj. 20') HW-GP-104 (proj. 20') HW-GP-103 (proj. 10') HW-GP-102 (proj. 26') HW-GP-105 (proj. 20') B1115 (proj. 20') HW-GP-104 (proj. 20') HW-GP-103 (proj. 10') HW-GP-102 (proj. 26') 160 ES9/1/95RA# ppm Upper steam zone Water table ppm Lower steam zone Screened interval 100-1,000 ppm >1,000 ppm Sample point Scale : Feet TPH in soil Soil Chemistry 14

15 LLNL Gasoline Cleanup Findings " Easy to build steam zone below water table " Rapid removal of free product, mostly as vapor " Electric heating of aquitards effective " Vadose zone extremely easy to clean " Increased biological activity # We should have measured CO 2 from in situ oxidation (physical or biological mechanisms) " Continued attenuation after heating ended " Cleanup of groundwater to MCL " Site closed three years after remediation start 15

16 Disappearance of DCE (and common sense) led us to investigate the slow oxidation of organics in water 10.0 Fastest Relative Rate (Log Scale) Carbon Tetrachloride TCE Pentachlorophenol MTBE Ethylbenzene Hydrous Pyrolysis/Oxidation Every chemical tested has been destroyed by to below MCL, at varying rates Carbon dioxide and chloride ion are the observed products PCB Napthalene PCE 0.0 Slowest x NDMA 16

17 Passive destruction is always part of an active removal scheme. Thermal Remediation: mobilizes organics for removal. Hydrous Pyrolysis/Oxidation: in situ destruction. In Situ Thermal Bio: in situ destruction Cleanup Rat e DUS Removal Biological Destruction HPO Destruction Time 17

18 Physical oxidation of dissolved naphthalene begins at 90ºC umol/g CO 2 umol/g O 2 70C 90C 100C 120C Naphthalene Destruction 10 C/C C 114 C C c once 0 n tratio n ( umo l s/g) o r p H C day Elapsed Time (days) Destruction of dissolved pole-treating chemicals by HPO 18

19 ) Complete oxidation of dissolved TCE: typically on the order of a few weeks T C E C o n c e n t r a t i o n ( p p b TCE Concentration vs Time In Presence of Excess Oxygen 90 C 80 C C Days Target For Beale Project 19

20 Conclusions Thermal methods can rapidly clean source areas, including NAPL and DNAPL. Rapid source removal can be extremely cost effective; may be the perfect complement to monitored natural attenuation. A variety of heating methods have been shown to be effective - the key is reaching boiling temperatures in soil. Vendors now have considerable experience. Applicable contaminants include VOCs, fuels, creosote and PAHs, and more recalcitrant organics. 20

21 Visalia - Large Scale Remediation of DNAPL Creosote and Related Compounds Craig Eaker Southern California Edison Co. 21

22 Southern California Edison Company Visalia Steam Remediation Project (VSRP) History 0 Former Wood Treatment Site 0 Superfund NPL Listing No RAP/ROD - $45M (npv) for Enhanced In-Situ Bio EISB would not work Superfund Process 0 Very High Benchmark ($45 M) Too Expensive EISB Wasn t Going Work (Especially GW) 0 We Needed an Alternative Process Cost Effective, Meets Project Goals A Great Recovery Mechanism 0 90% / 10 % Ratepayer and Shareholder Split 0 Insurance Recovery Thermal Made Sense 0 Cut Costs by ~50% 0 Provided Technical Solution 0 Goals were achievable 0 Manageable Timeframe 0 Reduced - Environmental Liability Book Value Implemented VSRP 0 Injected 700 M lbs. Steam 0 Extracted 1,400,000 lbs. (PAHs, PCP, Diesel, Dioxins, and Furans) 0 Accelerated Mass Removal by 3500 years 0 Thermal Treatment Cost $57/yd3 22

23 The Visalia Investment Was Driven By Favorable Cost Analysis Visalia Pole Yard Enhanced In-Situ Bio vs. Steam Injection $50 $40 Cross-Over 2006 $ Dollars $30 $20 $23 Million 10 Yrs. NPV $45 Million 30 Yrs. NPV $10 $ Year 23

24 Project Success Formula The Right Stuff 0 Environmentally Conscious Company 0 Pro-Active Management Willing to take an Educated Risk 0 State Directed Superfund Results Driven Enforcement 0 A Superior Team LLNL, UCB, SCE Engineering 0 No Entrenched Thinking or Culture A Can Do Attitude 24

25 Phase II Steam Injection Cross-Section 25

26 Visalia : Source cleanup of a major superfund groundwater site: 1,200,000 lb creosote removed A yield equivalent to 3500 years of pump-and-treat Prior to steam injection the removal rate was approximately 10 lb per week 204,000 lb Vapor Hydrocarbon Burned In Boilers 210,000 lb In Situ Destruction (Removed CO 2 ) 607,000 lb Free Product LNAPL & DNAPL 195,000 lb Dissolved Hydrocarbon Activated Carbon Filtration 26

27 Free product recovered at Visalia was an oil-in-water emulsion; this was a key aspect of DNAPL recovery. 27

28 Control is important Monitoring of heated zone; treat the whole zone! Vapor control - it moves! Suck it up. Hydraulic control - just like pump and treat Vapor extraction system at Visalia exchanged the vadose zone air once a day. 28

29 Visalia Steam Remediation Project Progress Report Groundwater Quality Concentration (ppb) EW-1 Groundwater Quality Pentachlorophenol Concentration EW-2 Groundwater Quality Pentachlorophenol Benzo(a)pyrene and pentachlorophenol are the regulatory drivers Concentration (ppb) 0 Jan-96 Dec-97 Feb-98 Apr-98 May-98 Jun-98 Sep-98 Dec-98 Jan-00 Sep-00 Sep-01 Date EW-3 Groundwater Quality 300 Pentachlorophenol Concentration (ppb) Jan-96 Feb-98 Apr-98 Jun-98 Jan-99 May-99 Aug-99 Dec-99 Sep-00 Dec-01 Date EW-4 Groundwater Quality Pentachlorophenol Only pentachlorophenol remains a concern. Levels are greatly reduced and continuing to drop in post-treatment phase Jan-96 Jan-98 Mar-98 Apr-98 May-98 Jul-98 Oct-98 Jan-99 Apr-99 Date May-00 May-01 Sep-01 Jan-02 0 Jan-96 Jun-98 Oct-98 May-99 Sep-99 Jan-00 Date Oct-00 Jun-01 Dec-01 29

30 Visalia Progress Groundwater Quality - Pentachlorophenol & Creosote Highest Recorded Concentrations Most Recent Concentrations Groundwater Gradient Before Treatment Current (4/02) PCP Creosote (ppb) (ppb) 3,890 2,700 ND 5 EW-2 1,100 63,000 ND ,600 ND 159 EW S ND ND ND ,400 ND >1 S-14 1,300 93,400 ND 2,550 EW-4 S-9 S-8 EW-3 ND ND S-7 30

31 Costs at Visalia $ Total Project Cost - $21.5 million 1996 through mid-2001 $ Unit Cost per Cubic Yard of Soil Treated $ Actual Costs $57 $ With Lessons Learned $38 $ Comparative Cost per Gallon of Creosote Removed $ Pump and Treat $26,000 $ SER $130 $ Estimated Time to Remove 1.2 Million Pounds of Creosote $ Pump and Treat 3,250 years $ SER 3 years 31